From 5322f5db5b6e567cad0b7f922cefed72aef4e814 Mon Sep 17 00:00:00 2001 From: Morten Hjorth-Jensen Date: Thu, 26 Aug 2021 10:16:05 +0200 Subject: [PATCH] buidling up book --- .../_build/.doctrees/chapter1.doctree | Bin 432356 -> 440936 bytes .../_build/.doctrees/environment.pickle | Bin 186462 -> 184642 bytes .../_build/html/_images/chapter1_11_0.png | Bin 0 -> 6285 bytes .../_build/html/_images/chapter1_13_1.png | Bin 0 -> 9553 bytes .../_build/html/_images/chapter1_27_0.png | Bin 0 -> 13856 bytes .../_build/html/_images/chapter1_3_0.png | Bin 0 -> 10016 bytes .../_build/html/_sources/chapter1.ipynb | 1334 +++++---------- doc/LectureNotes/_build/html/chapter1.html | 1088 +++++------- doc/LectureNotes/_build/html/genindex.html | 2 +- doc/LectureNotes/_build/html/intro.html | 2 +- doc/LectureNotes/_build/html/objects.inv | Bin 753 -> 742 bytes doc/LectureNotes/_build/html/search.html | 2 +- doc/LectureNotes/_build/html/searchindex.js | 2 +- .../_build/jupyter_execute/chapter1.ipynb | 1501 ++++++----------- .../_build/jupyter_execute/chapter1.py | 827 ++++----- .../_build/jupyter_execute/chapter1_11_0.png | Bin 0 -> 6285 bytes .../_build/jupyter_execute/chapter1_13_1.png | Bin 0 -> 9553 bytes .../_build/jupyter_execute/chapter1_27_0.png | Bin 0 -> 13856 bytes .../_build/jupyter_execute/chapter1_3_0.png | Bin 0 -> 10016 bytes 19 files changed, 1632 insertions(+), 3126 deletions(-) create mode 100644 doc/LectureNotes/_build/html/_images/chapter1_11_0.png create mode 100644 doc/LectureNotes/_build/html/_images/chapter1_13_1.png create mode 100644 doc/LectureNotes/_build/html/_images/chapter1_27_0.png create mode 100644 doc/LectureNotes/_build/html/_images/chapter1_3_0.png create mode 100644 doc/LectureNotes/_build/jupyter_execute/chapter1_11_0.png create mode 100644 doc/LectureNotes/_build/jupyter_execute/chapter1_13_1.png create mode 100644 doc/LectureNotes/_build/jupyter_execute/chapter1_27_0.png create mode 100644 doc/LectureNotes/_build/jupyter_execute/chapter1_3_0.png diff --git a/doc/LectureNotes/_build/.doctrees/chapter1.doctree b/doc/LectureNotes/_build/.doctrees/chapter1.doctree index 56d7d996c6cee4f6aaebe76cb3d8a92aa02331f8..5a49a4a82229ef9d965f69f78448a3e6c027a875 100644 GIT binary patch literal 440936 zcmeFadCV(Wb|2QQncg@(Qfp?^@~AC`8a-*=dwlP4-+Gpw;=b>DPp45Ni)4|tkX0;h z^1w2LAQJK^5$q=;u#Lb-;3P;Q3rP^bc9wyIATb~aNB|?rAA#aHih;mD{t6NVO6>gZ zttzs}uim$jPva3JX!?Dux^?e4_nv$9d+x8k_1kZ}@#Y)k|E_*Yolkga^MV}?C3(oH zFR$Jm$^%X0q~*)M{PM|PeEEZykFVZS*tIHa%7A-$^#)oDCcH3IIO*kQFRy-*G+*+9 z`oj4B9hDm_d0EluP@5lq^x45o;YOs5cJi3z!p4;Yu>Vat|i#e90-^7yNuH^{Q>kDNv?pFN%|mvi-Fw|l)_ztDIYQ}0%n+K}fIHz!K? zugESXcIh5%Rd=dee_n6o{oY{EJ@I+mBU#}DP9>nJE7CKX8yz#+*ynvA&I6yUl;!3- zE1p|FXVvFxu0LnT+!J$Q-oqSmm~xPVNg|WCu0A*gQ*ZvE&5Ql&*}TJS1SW4zzA$-O znmj3CMJIn2%Q*RSAPFb__YM5-5&rkfSjM*jhof&L(Y^8E$;178PAY)n#z`X( z{1hPgSrg_wNbmW6upgE+#mR~D39Is|b2(9DZ9I{+CH};35IoPRgc1dvo>k{#Rb&Oh z36hpTAx@_Q&1A``NkeRoB`Xayo(J9$-9Kg}{+6zi^k0&2qefK;>Uhb## z?4Gb9-?tJpHu@L9Ppw<$op1LQr~A7ObI0s|@1Xz({b(|_c5~Cdz!t|$zr6b5WGMiJTX=Xpm7w`ELJMNXt%;XF|zu%VzXV~w}&fogN8*gOfHMioFXU_Em9N0OR zAu;;`=Un6edF(@D>+U$RbsW4nmlbTn0Y~<%rjotLu2^1R$v3JzTCO2FUpU+Nv1App z%QY1pEhkRai5)#upUaYiY_z33e?FAgl9Qd!6&bq@PpQpsEO1;vTmu;Ls-b3e}nJcB>=xiK#hY5L59 zt5e0?3!Gyv3tLGRc~&syRRMUi=#*q0JwWN=hIIPr;u$zQE8FD+^KPz>6FYYX`Rn|& zKHz8k@;TKkbR_^ffT};FEbxV_jG+mT<#j(HGvPQ)-HKIs8OwXhGp%z&7j?!tN4(CQ zF%SA8kx2#T81hsKVOBsgqKw%?3x#7xP-iGbkeS7hc!oui#*P6CmCLDWAQ)EFyBaCL zUuCyWZZn5GA(|${NS_nr^?rHb4qt`#MN`P0J36uEM zzd-Ef8&rtBh5HMpR!W5jH4?r^)C&uS`tNH<*{Jh|0J1Rc@0_KjSb#XQqlZENJ*x)s z5G~1q!1;YW42R{9LQ%8e{i7NGVRQBI!}UuG#^?0%Fu1>cj&f^({Jef11pDOeJ-?$K zfny!^J$PkdAWCDaPvwLy5|=XEW=D*5-{;Bm$w$*CaF5@OjlyuBev`C*QncZcZ@#ch z9DbZ)f4f6q-)a7uL*Qc!tJC7aJ|#{7>feEaMopW!roc-;9ROo)NsTgUk`u!-r755Z zMTFJ%j5OZD79?gE`Ba9fw1nA4Ug90hrsZ0O?WMpMN3D4l=B_NkdjJr&y7$5QbzdRg z23Sa{EK=hhj%`sS&xr4jAMpbR)O3EJ8|+J2hJ_46Zp{G!{HspnHE9f};N@7uyo3kC zG+}iE+367Yz5=TjCLl4xuA6L7Q-Q0!!{A_?`GEi?=w5SZS=9%94ibx>keH3FX4fWp zk=3)nf1wWVjgdwEGiTG11)mFab`CC|oJrLdOwR7=96Y{r-sU>+)-yY@!QpnM={5D$ z>r69)ZI^+63NBQ`XDYJ11b5H~aZDP!0H1T1Jj(FJnJze;(G-oS1NFm!Nj-B~#R_%1 zH4qb&l@Vd4k5!Iy>h3MVr4UJ+sZMS=c=E#OEP}bhiw1u#;d2KL1lR4}?3( zz;HXG4xcdU!xvZYn{-2yNmw)bf9<)E4L|~D4r^fh4x#n0w1!%3 zBnqN!MwoAjxQ!7jN32antOWWyG;qS`6)56*M-V$yV@I;j0Ese)zClPqD$oFuG0F^) zt3$NMNkcY7BoMK(SJC%kMK>3atfR04Os5GALUIZ`*U-O-SYrDmjbJ`O?4D!*K#UBK z?nB2TvX3SV2wNhh1SA^~!6wS&OGrXIQ4s#3G6}NFh{>U2D(L$B`0xq<&gUdZG*Y;Q zhR_lOJ1;pf68=UM4nOf9DlhptN$SBYQ|t%~fPcu7F##VWI+^un&IUJ_NU|W0x6d3! z7R_i5gARC7Lgp-g@705hONFf|KGm+J8xtYYW$SPMp%=bn4H1)SSBy;2*QyqvTi3#!+`I*#r zOB@CyeFvE_lDxzsJBUYu_)WPXlp|K;1s+b;CGqhhG3**(gfdt`lvVOAP=uyRlt%&p zukuSEMSZinvw|Q=n{t~0^d3x<4h+cxa>{d(uBPb<(K82fcF-JoMbt4(6N^AfGA|LO zmn4!xcN2)NWhl1K4DAxiBlJIBB6)Ami26tdoh$qb7?D@>r`87& z?Mk7PK4C##B+-r)93-(C(9vE+>J2e|uP3sexz^{9G&BLF4MBK>Fra1jk@z$unXnj* zR~*JtpHcyUvtLtCi>~gPYAAB`ILTqG0Med_jyl7#3)~Q@u?@i-6vwYz-@XF*qjk{3~dEGS*5!*g)@-nj~y1)cNzItP=)B{V||Hkaw| z9E?(zQBy5AU8=`(uzGect@&Vo-|ssc2yZvN^EaB_e+^DmGcrNF^VG0}p#sylPfbOl z`0H8%?4-RJM09Z2GXnv6KhzY2g$H8V}W2_o6v;`b6)Nl$0wk3@mkhvg;BHjxR zy%n+-2XvUcw*$MNGWRAgNQ)C#(^h^hAH`s%^T!a*e`kykwx)eulbXpMR5iZ!{tN)Fy3L=0CW{EQF^dN7&@ZW1-_bJSC~A+BH4 zkJu9LGEG7vNvtHA?L-{|7F-wIAfS)Fz}~qHWSbKL3GvwhnEn26`Fg@bV331=2!7)j z6P1kc7vhAY;cF&NL)O{HL033!N!bVjvhI&V_?h}a30OF`O5g&xYxqZr>>6_W%rP2k zrrvIHS{4A~Os1$iT=%f(S+ji&)ij*+#3eGX5NSqgtkr!;M+Tl9sz8j&#g*@@$Ih|UCgj+ub7ctT>* zx@#Eopz`AhP$u0R+>MZ>^vJFr(}$!pGJ2ir7A|KIj#vTbJ6Y37?oecMM423|#Fxfk zL_nIKUGpjLAwCG%#3jhg!Wk~pAGXv#x`Uv#X_LAP#al4CRFC(@$s%w8!zDK0F;*Ap z@f@r^emFm{U~*0`=Rlt{_t{x2$%4p*Iy?ubPtIhJ7HrP$>Kr^$mtoWvbS~B7Sy)|0 zZdkFpP><(e^{sQn*g=Cks~a2mZ&%^^H>z;|!xvxSWgsOLmI^1rhv0+?HPNHf&xoR- z&WXMQrHE)PNE%0_2bku%c7y5&s0cyIKA36(65x?hj!I2LS%R_#jfyB~3i9HI+yG7i zlGcuFK@vMReG>zAuB(G2H;yVBG|D=X$pIx@Q85$=y?_Lq6bQ~zOP}Z}OOmz6lMpx! zKhxOI&U6`&$|lJDBbkOqULj)0Q0$(O3MDiJal%2h1KMzN9GoQ24$cWvlO;o0`~d;*?@Bt{qg`hYkZcP;^17I>TihS`yVf(^ak>?t~*cW5jpEPg~of6s^nQr5}lO9Mv zaWxRHiNmZ6Le=KmD}y|9eA?I_P&lq8h7pZ}3$&KV6Ev-C8!|FU_ubC7h?x&2`g@Al zr+Gxm)k8-!^792RyPv@MBM1&8V-g}dl%CNgI5>vJmgt2O#o7q;q$9^(A=SS$NzE`KV9j8mSnLuQ8G4Nsd_;2;kU%ii zPzwcaGD_OdkmbOUYB5IX5ru=wtIT@1lLS~bEjJ-?JUqN|5jl(fIm~WbZg+QtY$&*0mvI3iq z7%qlgMr<_tY`-M{619rV(Cf34x?dDlea4D#`*Dsqs|ctYWNuU#BW4~AN(1Q|nW|bC z6}4ebdIQKLrG{k3yka>_AD?j?=3qS7`_%5m;4H%E61{hbY&4FBg(*yLp@nh2&&Y;ssy)a1=) zmU+UDOJMIfs9GlO)<*}-txqxVZ{G*+kE|6?Pjt9_6_?v_p8ggY)GC@R?E17WqeX!2 zTOAht9Ybw2ao+zvAiRHs5r5}C*nZWhTVvHkZ^7~{sxnzt*LZM*tWkpUfW%u=u{BXJ zejN}nUw<_HFYg2XPn(Oi*OPApo~qGS;PnS~z+W%8W4BHpUVZx8Xi*+6%@ zKX2baE>J{#uZLLSd?D_XXHMi|(YP=@VglDLV({>TEoBEO z_(Y1MkV;S-37?a4$j+?fe5z`2&mcn!j##R{!%g*+xO+*6m^$7NK_)euNT``4Y?0VE zR2CS=0#49?Ob@P9)X^oCN60f8`bJP5CEZCp{DrfuJNRV0R2;->`^@=NlWYg48CSX$ zPF-juo===-W=Nk5PTafH38zkY8ZW;d(kF>yI4T4W+acNwA|TM|vF#U*Iz8e<&I)P7 zk4Tcph_i|~&1j+_i`J-;WlkC4|L2&2=hP$j zjQ9eHLzW(?Q#cd>hx3%I&K!F2fWoa9*m%bFg|_=TR1@%qKRzPIFkAeaY9K zvE^$fsl4LtfmmMA0K+aWM=rml^Qc?6tOff!eqb{>cE=5E2`O<2o^Ap8GX0%{QRJZl z#)8Ge`*<+q$K?E>y{t3o`BkqE1pTTOfbXwej-sF4NA_Cg_kQ0!cuuhkk{1#97SJ!$ z;W;?{;$75_11a38xeYwGE7E(ZXs1YD{ThSw@FaeZ9803fnp!X}SGlH5>n9kX!jX13 z*noUgdeGc~wBy@0Z9X}7S`O54c2_o7+)g>aY3frJ%6aeVOMOj%f%Cb5%wvld zc)0%VJ@V#rWH^Y|UR%n5y$iE0PSTF|Spy$C%anegrNe9+B6(WBegxD;bXk}y`k8zE ziaR|IXDq`;&?(YM!uZt8V$l#)A7P>(1%r$*#vvS|rX;cXknuwEdO-ImiECV$NAfAm z;*>^@7@w59*g65sm?Q!1^YKv9m)vwmQ>KYJ^xOFWFKcI#> za*fP8822wAQ0Q3*%f=_7XsO`pKu8^PAbszPEX{TzGKqRYbh(9!EE>sQGNjdt6_7z^ z?!g~t3=Z<*$9Wk(n!fqgvB%?f`uYPW?Vp`8!d+Tu`@xh zYCM%yV(;#)LS9)*P%K~ZL2Kk`G|Z^VB`Ja32V{)`UvE8K{5A^u2O z!ING$2O=l+FriV|TGM+%{&4{HK(Rj#(1P`T{obVc(j?6fjZg2$diUQQNb^RGZTdf- zuxyLcw!dES>Q{-M)6C<7FMB{rW9X^Jdag2cg&Wn~_k@ddXcRf!C$S>a)U4TnF37<+ zo!Tv`4V1MTsFUdWsmnlOQTH#^BbGUEvdnfGzfGIy!>8r26VB+w2J#zehUx}?#U#)J z!A$?*H5vc-X1V_l$!d;xR8)f?#~s(Wr5Q)=pqNXfj2Wp&ODnh&hDi#uKR!C1h=cye z+CJE%e%B;qGY`p4PU2II09n(zC`VoSA_-&A6ODv@Qp=L?P=0_6UqnQ3(I;V%^g7U@ zadf?41{Fd;Y;t(C?j4aCiE_gwW)KB9X{oHocV_-I-? z*$g*IR!UO;x=B*y`|-T}cZL0THAd}W->`51goz{RD6lyJ$x2xQ5SAd_iGf^m*4;Rl$b zhiyPAQsexs5$Ym$sp8678ckdiyLC;^z=BGRn!9?Ci^V-L=4Gp>TZqFGM1RR~(0^zFpodprGAb32Xp+v3JO zXll3n@9wyruKsI6{`BC|F%t*VmY);%1&HwJ(*derx>0sTefsoZZ;_p#pW4753a`%g zubv=oOAc1ya~UUOcYNfoS6Ee{;e(?_YSQLmmH3vid)sq#RJ_Z{yP2#Kf_VZ%~ z0xVDV)Ns`7qv>y)bb0go8dq`sb(1bkfM~q)uk2@@vZsG;7Uh8YKd%tlEbZ;;(|!?t zaKFvD|L(9kZye0j@AoqDpsrc`FMp5GGf!}~Pgj88vNRlz*niiqfUur1GS3x}Luneh z-P*G0At)f8cDg2p4KB~IXU*S-JV;1CkJtNUDYBrb$1kDM$ zbNl_gP5*l`MITN70sgtgO)w#sto&Q%gFf7 z?hAoLsI6pR`ji}Hq#_%}8I)vl%ZusurJ)xDScp`o=Q@dfI7jhoqZkXVYl8uC^7`cU z3r86!ams{vc*#(fcqE`Yf_z_~N7sdCB=L?Y##<3whm_5sWx(}=ne?DBboNKgBcfZF z{Xp-b{=7%AN3_Vjf zamWxu8NU#r1^(nF8Jt^%DNP#@5D%KpM>j0Cg*V;U421WO2>rb|`NHI>eoEw+6?}99 z(U?SkW1BspU)|UYfc%paK<>mV{-(JdOf)F3c-0-w*O0Yy@|ZoD6EgYmG$rn?Y>J|KPlRCs;}8d6f4B`?`=?+eu824ddr)!J@By;;OmzI{vAUw z9!9v@Ke`RXX_LUA-(C%9k0*F^IT*icGr@<^u54f1zJ|2sk+JF6ACVyz$G(sa64SEe!lskjN~Fy}vDb6l z95gu-S|CnXVQ~G>6sdlVD6mmK+EVA~-~iIi8GPcPrL9m96WZDyn*A{)dL&1$C$tZ^ zA?BtnWe8r_S|b*Rmrxg*Ucd!hB_oc9KY={$53Tmy!2lOol+`|9L8>GGPF zW7-5fMNKB!6X4YgiISzW+yi<*f(Css!F&8ja7ICjt;57mN8d8|!T$TCq31F_lw>R$ z@a(RzTmSH6&)M+?n>NgzR31vW7CbJ{*}XBb*eTyTcMDliIlr%SFiBoyFI&*LP>1K> z^ld_5*iR4QEEZJG>*pK{;ufw$_|ikzn+1`JbaxIu%-O8gg3SdwI|rB8`Apw}%q4m| z2b-U}LH_JbwX2(2*^GX>n)J6!Z=S81^y_eN+i5cpz5A3SxKf>;T4UGdec4v`{h^>E zC%stnGaePfd6w3FB){Aim;dKD*{pl8Q85y!UXay9nzKAjIx%O-BA;o_5=~LE`tDvh z+vL8;fu=%HBW2f*E)1phNO8qD@rlD3eyB_czcJFBjY~x=G2Tidg|rjMe<%KeE<-B5 zJ7~psC=fWMfd3f!k;TVvfy{JPlSn!V)2oJ#1phLS)V zgBJ4WE-PZ8?7OuG^?d7{W8dcFP`wC~bciV0EwJhQxV_a~|{{n5B-FS<}jFaPU`OU8`ti(4=2Yf8* zSLbzYEen>fZ;W#=k6&Q#Tadg+cQ*Lk-GEo`yw6B20K>!jiY?=3U+R1XVi7UCuaecVlFowO;G(utvb|7Q+9Twjt11Jv&tLrzg3?-Eo0D+aF-MP;5TA?gZ{c8~Xtab2%0+8> z+})^wf;MXw3D%J8R436BQAu|56V#%Cw*nJMs*I!Jw4O!xsQ8Fhi~-)Xj5A4lhS8w1 zxFs8o6Fubq==*(H7$Po?e}1L&1LyalKzzFS-22SA`P}#H`$HMNsLki4&ny!E`>zub z%O-oh<6uqy&dDaT^H|ul`KrxaTfnw`Z8O>tGrW4UXi_Bp+EX9pSHE`x=BsrM3orb8 zCLNjxQ1#&(w-dx+@TEeK3OE_;eMng{FGn- zNCG$7P+zLCm+`{Pg4cIEYN(?;I!TblT}QMmIy|0=k)OUsK_t=_@^%gD z_In2+*l)ip%R`&KZ{I2ZmC35?c1kAYNq0kUb>-MuZOI0J+o@*P*KDU zNmQf^9JvUM9O#*A3P~A9K%`9f7cB^&mo*}mt}Zf(C9R;i7%4D7Dw^rpaSoCmtJfu< z#y{P2*hdzZzKGK3I6I`0dkz#?G}7Nm$Fwm@#&P3}kq}n0Pim!hff<-QL^tWdMiS(haKaq! zK%z+?NJ?1*Ii5+&F(9JJP$^F+44;zIjwp%%vyz^6M6$Azh%A)dK|xd`h2h$3-8~I`VThBK(h?GN ze^Pg=huF#J^rR$7Hy~Y@*ZCO~NkWu%9?2il8y+gwfgQaO# z?SCH#c>f5U^-nsx*~m1CZ~nI?S(>0wBlB&Ycq~YpPi%m;Q>e|}Ok)de+OLTcEkM4u zA?}T}MOO&#_v!t1iUqF+cVPqfStR!Bmm{$+>m+6{01L7wFKke_ncyr_j}R=exX)oL z7Qimin+-O{3-Z2|jhP$&_~BvqA3sOmfAezdW9LFS-PN~Cy*+!eH4NWGy$WL2m=~>) zI}vHs2)bo>^<(`!1Ja9716n_Utde?%L=2d25;Qf~BuO5lCmbQD?0C&5&V~_DC7PIS zEmGM@4>uV_f8o-BKZl-7hFpBXy?93QxW*c)4ZvI@kzYx>@&=YnKfF-pP^)L5p zZYomLy@797LIt!SC_5Nvm|N&;utc_qIUqH`GL`x;#-c+5#ATwics(v^-x|}SdpPxE zEuA9t!LRR(gQiw&kR^hmNealH%?d1x`@$WA&;A!DeAZ4P*|e#@2KHpZ^R;w;Z)`2{ zw{mILX+ie#eV>DA{Wb7(3$m}J`*ZNEUR>T-uzd}kpM&e4yIB%i0KcuJ4HU=cPT!(j z(p3A5Uk829eEre%|2$zpH`{tf;gSE*wDnBZq~{}953{^U&$5{BP3!F_&N=;k5>PQt zeVqn6-+k}cYkV_YM$(5pTu<+=RalnbK3~}^$IXn~!rJl|RAemRU$onE@GD|TpO!F&Q`n$~=(g_T zU(3o&uct!ndQ>JvtN`JB(`W^W6N^%>6ndJZhNKmVbBJ)>=rInO1Y?YulundMIXZX~V`{GM9~&K7Cgev3!a4IM{ITw%1H1j>9oJ@tZs)-N*feqM zIPle9$dD@_WhX~24*;Jta(oQe6ul~hIQ9s8rcN<9^2tqaaePEzYhfI3eC8~BmKv0u~aze1Q43R#};svGS#T7~VO zT%YN~@?3q5M1G(1>T~uvDGd5tw`&in<3D5O>K|uxUrQ9<=VBic#Vw=!8FLilbo_Jk zAhIP%{p;pPrzhw?L8C{Xy`22=XY{}hNlX8pfz#xDBir+avT{M0u3x&-=N3+N#}BZU{~p#->mNH7-nW8b`WKEL?rGxc z_&Iv~k7_z9^H_vTouKY8j6v5M{b$vYwB2@^8v zcS>Aj?hmkDrazncyO~x=qWB&`LaV2V;+=m6y}wN}SE%xeoGubL)JfG@+2T$Sa?{9= zNFx$fL?{c_=t2T$eB||b9wAbQ03*qybu`e2q=|hLQa2=*@qzQv_W|^;Y<}>Fc1UBH z4k+6QgbtA&Ykp%Cdjz1ge;N^`+v(`sD}84S>UgyIoYzmOc%FGbd*pmf^>;m|jaUU<#R|yWl^5xVB)KvU7U5 zZPKP6kU25Jv8U4Nw=Oqd?-;6&1+D!f8yN4*mE8ALrS=qYb?;AY;JjVA-wSVB<-VZT z{d1O(zX#{GabYG@M3Nv#6ZxSU>N+WO!>3R6XGD8Ob7-E@{OMDi?}R!^QtYWKi71ig zq)^4&Jaz`3yqQ0Rw8LvkHc81hZhO(452Pa#A`C6fU+CS^e;z-1`V_f+rpaL@?jx&? zm?|(r=t*_mS}>q7O%0e0&L+K~OUD|Gb#2>W)B-fGu7K67DDe4HrJ6uvjDjiA%?IKg z;{=L2!Xo-$yyPIJJgM%7aoR*5C!ZcqFwh}DNdyZ5EXiRc0RRn7JsnzyK7wKPkRJ2` zLr01^*fu$tKy3)vJE8{5h#Vsz2E>k2kQ0ZB`V?(q0zCRY?j#o3f6@1d9@9Y^1T}DS zAMQWHB|srHEE1qb>@}L&N&JVVQGsi>_eKQ>V81}5{pJzne#Alav~dmTS5cK^OWx-z z?1umFgg+d+3chlkSWV_^_v)}Q?cQRvzk4#;%|?LXPf6{kg?5`=fhJl3_#JZl9|srhvJXrCZRs$454( z-g!E|d!7)n0DNvwHdx$i!G88j5AYS(FW490^srz?f7*a|=Slv>JBV1zcz0@U1J5zP zc*hXgCY#v*K5&QqBXs7!*Yw>aF0MYb*AXljzN&={Qg@zY@4ZzJ3qTL}*ap@UdWD8# zPW~6j-}{NDzOBl=Cj3SQ^Ofs*)M2nIySFyL-+7|n+pB1nK`ak#`aYhRcTLUCoSnmi z<0%p@H$~sPqY`OB>CT_rhVbFU9nGAmJ)ta?s z)SFt_jDAEYHyYc2^+NM@yGgp25f+p5y(^?wEhnG%1$i*DUV`$0W9$5m)3 z5~wDPUw?5->%R+pZmO4d^|mCh6?XpeH}DrVR9}`R>&b?cu=*V`7f*`cC>hGT1GX8e@9dcw*Sm3iYI*t66*jR^XBshTN}gf8hL%@I0!gZwDRi8s19)fTM5{#qn0@T5xive3#BXz0F+v1L3 z*Bu-r$%f#3h>N|g(}881noqDLe|WO7j<)30+dA!-RBiq|rscnX(#6fDry<&WnnArL z05tyOyFZay(0a(%HdA276HO@1QyXaWHM`wmfyVy5&3MO(-22CA0lFMHez>3htK;YB z`w#E67KUDDzZPF5ly|bTE$G?4w!!Y^**uumr*4PS89$z-4^A13Nut+Z9A@kKJHY$T zE;nPpbixhnd7Nbq@A!eutR8REKYJ6~G-u}KM*CT!+Ui%|e}HvmPexy}d*O8B*7%y8q_N^;``!kpyKauFx2PDi3}HO5>GPO0 ze)^i^GbhaE&3@)65ybB`VK?%1NPjrk?<8$4INs+gn~AvV26+ z?z*+EJVZ&lm)5$UiU)ja1L^Gs-(GCl+EM>9wn+-SYc9yT1`nkOAxCOqoR|rp!bT*` z1ipW`LF4d@-k*UaxCxIdC5)Jsin^P9fz*`6X&+Lgn&#$>_%TY)ArhG4IFwsaNNE%# z)^E@`BA>)QgN1=h&rn~?@#z@$4ySBiIim62jVqqH&5OxW6rRw?pq{&nL9&uj;6#5y zv_66b<<_#oqY5`-c9yss>@aGv(D)l#!`y;71gIPqL+(%-{XT>_lRDJOJ;yw zA}mXgy%XeE43w_`93lPxno@_olQoq*baQ6mr6c|Kc z?s4m*19e-UV%Y!eWY||5QMftdXU&^44uetZ+TV`xEz`O(A*I&UM>i0kWwLH;cJCQ7 zWhmy+{{RBF!1bWd9}e$fP2Vy0kh!4y--p%QKf-u_aG#}o@`|OrzHM(=+z(E7wPoh6 zzkB$!e0cRtuTOC|c42|ye&5+ZcqH*}bQk`=cU}TZO?dlFbAzBaH<#%YIEL>gH}Fq> z`2qHoJ$ye!xUYo^TPFP4E1NmJnY~#y@0Tpx-7?0}W1IeuRp)n2sbFrU!-K<;fmmYAt=|UVs|9r9Gg|+k~VAjoaJW zAbICS|2b-Q+Dqd-`d2r#vKjq$m25vPb}HG`&s$uy^lXQIa9KrNUxm}P1R!`zt3C8n zmMk1j6yIn4=qbIpDP#{&UVT~Is>{!%{tF3*PSlrI@5Z4YmGHMEY5;P9&G>jKV#KR2 zlLlD{$yK(kChuK+89E_{vR&7r51s!By6G<+Q~-pr$$>eC;}@M-njJ?nqe%1)XSs1R zEcIopv_KweuEF(hCdrwP4n4h?Z>b6h&N3I44tiK#ozVk;FZ45cpX&$l4ENxvGtS}I zAS`l{!GhFygh=#&$(WW96GjG&9<6_m8&bBiunFaA;(lP}Grh9^xd% zm#96rer#sz{^>gZNnhtZGoq4Y2flDqY>>nOuih0{#DWl=0*~S6^q`xcXo%B|&d#(w zJ2PeYu|;v^7=ZEM`bKab9M=5?4(lEc>`qnk^^;=kuNy&{xzQVrzEZCBlZ0!PwdJeW zmfq@XhgLBZx)!)27I*dT=P0y}9@Lkgy}bGvsZWw_*(EsVlIx=wCgGqjuih^iKM&tN zzWS2>;pfZkoHT)tVJ@)S&xx?Ybl^}qOaGky8vy&sUwrxa>H}Jh@^f;se5}ZT<>l2| z0=f0&C9db-1x`XxPJQ_U{K8$=W6nL7#xJi}cM}&bsBUpGB{z&q?gmFM3Lm?R`i0^2 zoGX~Ddngaw1^pm80atZ@-q#SvbWb&`45xg~ZMXrlXWh62V~*e6&;6f!UXU?Vjg0v@ zOn7+>B7gboYX9nsC$Kw~u=d?R5LrrBzwqr8bq?T&wBRG+I_ zW%9{f$rGy$;>lSqR_>)@y;d%k+O#rEG3$#LvkjqJo29#h1T#)fJgHPI?O&x^DIuP& zrDB^wI@_EL2c=}a=BYE|TrE54#hF^JI-6!|^S%~#uQFP@)f%#k1sh)ZB}MVO{36Fm zi(qTSaa>4QxXZzw5=f zvz^aZ6VrIt@O6E$YAn0sd%@;%GY=QTMK;}Eg;znh6wUTyg~hPiO^4QlHCGH*5`kH; zqcz9noHC1qrCQb%t7e;OxELL0*QrQ+$W*$qBBL%ELCdd?LWsjic+=4QbRc2k>l>vrSoJGotT3wa;<<73X-H9E4 zZ&wPmyzwblQi>zdD~$pLDI-QjrMK#(ir(d_$FE9C-rKFJJGJX$eV*dl7h7rlW?9Om zlG19^Ood88r8f$Syzm-Zd&4S)+sN?DRC&cs>)%j%*SUd48~`8E>W75ig%U5gZo4PKJmwXB=WYkkdx#-4(dN)-Do zwWn0Z^Kns*Pnp)bAjkYnxtd5bv2i6(-VTDB0MkuO2dPmy$dr^)xRlV!yZWwPO=z(W zQyXVI_@XtQ&L!`{-A@jrb#Wc5GRbv(kYcN&p41JE?olz6DC=V?c$*p1K47@!D<`GyTx@PlELFaM%s0&33pfQjuNfl zHW`R-7zXnm&kfvC&tH-gjj>i)C0tC0Ne4=vm6~D_5wE(*#MFMI6o}R9)ipC&xSkN} z{dFY({sFu~%&^_{Z6Z5g!JXbJ!BRwYRqISvjk&QA2J?KMNvz}DRNfzqMze$QC|8Lv zt(G`w`N~5Nli_Nk#6lK_>x_4|@eb2oDb6GUiD0&0?+4<6%qZKLw?>I_T*;*(Gp)A5 zGR2C!QZ-*M26$J!pYzx2ZiZz>KWLsm5lSw2SoJnMi&d zOV0Y+bg<~-3+_#bVFwwXx{gIAT&4h{Up5nVBV;aj+MrIVp3s4CD3JI+5`>vCOZ^^^#O_`BUymcu)+F`oYXV zTvhxFG2!<2n~av2%iiUh$k1XRM!QO?{39$TFsB*5pOjn&L&$nvYBKfax%UP`8wVz zqlKh?k*U{5iHe+1g=h0)5P%AcmEn%SsbLO|NrfY*a6k zS%A@j-Q`ztH#f+M;;Ow)v=h5(d0pnF>r6W)t_B-*#dH!wEg2FP5x?q7iKVf^^csG3 zG+~)(aMh~?O3`Rz)M+=|?5@)FtI={|Tv#X8F*Y#g^Icz0>_sqRLm0<-7%GG#!}39R`jkj9d8BvzO6Cwb-I{n4Tp-`?_%;|qP5PjawisUmn-pV zm`Sa(fl4H~^SaWRS)XT8eF)bbU)0G%Q1))g;&DLY9h`fOaOaMS@-+yNpvvxH)`3wXV!|8 zyd5c0D@gn@CNU*{hKt16@VGOsbljQdV4T$Af+rN?TrNM88z%xqnIEhNS}Y8flF2fS zNX#oOqy0c%EX+M?|9qYXrzv)N>sB=3%_fRrHMvgI>zhb>n3~NvZ*`Kaq-$apQ}1r~ z+bL%2U)2<5?N3YfV0J#6h&jm1x-u5|WF)4|7T#JW#eu}yDbcgsB(w1n6AL6GIZq(D zZRAo-snLuv+4fA-5^nD}%jfFJMzh3s`B^7b^^PZyn|aQ&N#wkN{D^g>(v{h^6i~)|cA4LKm{P2mEP3kXo}Y=2VvXUt zmaHW!wL~g1%tXNE2a0FaO-GXD`U)RuP{W#Xzsh)$>(0oZh0K{|`9jMxiq(?6bhEi- zs>$Xo2sNoT>(!uxF|~SQmSQI9nvg1vld)8;Ty4hx#I?<;YBD$JL22rD)4iR%TZnvi z>M0aucHL_-9eKCr=-l+v-4PSjKX2+V4N|1e17nc1V({c}eGMk*-fsW?s8K242iXwRMG_aVBtDAyXhJEDm zGV>(fN%UQOT;=Pv#7-7Fbr=WPS}r>d&N_wdBI`<|Bd|-xv+7`55LQju6F{-1NsH;N z)HQemL>l-g?N@1M-K_~%^kz46U%Jw-`<6<6DsL-VFFVPu8da~>n=?ZQjFxxg2{mWq zVsKeel{V`Q&C}U zrp;K=Gh2cz;&P2=c$V#W#ZjdaJm|=R$YtnWHzp^b0dD&aL2~J;-D4j_B2l&`!rOqtc#lvjGZ`> zR%4Z9e;cY6Dq=e*W*{rnc6g}qd_7msM)Qqkrn)LllWJuU>nxKom@RQFHX*#aliH~q z3-aK0awt1iJNTkn83dErgyQ#2R>N?l;|ulK#ugtX!a46^&ZT|S)39^-q3V04Tt{06 z32kI1juo-#OrRW#7(^w@<^Oo=+#5p6p_rg9Vdj664yslwVJ`DqpJ0 z-k7Jb%Wkp)KMloa)tIa2V+&yT39o0cF2RWAXX9mE^~F4uZMKGH-V(zzxk|hOsJ&!E zcROLGk=tdH^W`kS@-=RloEKunxyw~dv*Fx4EcH^OdQcjqh2>02@P+M8i|rPLShu$5 z?mE(-$!E2!|Fy`9u9%;@hr#ePSrtpdv?ujusVszIQ&e{e-?$a>#Osye&Lve><5@CT z>gL>xKc0cjJ=!H%2z$t_`lO-+y3?Ysly=MB2vZsPr!i?Wj}J4`K$-~_inGR=A%1|> zuDlfkrEnrJFMCE2DcRoyM~UVr9dN4=A==7x)7^9+9%0gUXVZFlSskvX z3s)w)%kAc2O&e^3m2jmp3@l)Or`Mx#tkw(2b1_=46xJL+%r8R=wV5h6^4@x=xGsmZ zMXoa4RQ$+T7LtBCCSdq-hx-1EW_zp$tQZ3 z)r=NitkYes=U#?~f#xJFO7V6pH_LZdZMN)J$;~h!eN@^GRO|=RCXDZtcPE%_!N7@O~ns=IACTn<= zo<+K{C&F^g$znOo1$IKN8V--afAis8F45d2vO}?+o>wzDB`wY(y{(X!1HI&I%FLFn z+`LlelAaLwMQbq|48=?-7uu!rEhZZ)v}dXM$me3biPR8&pLlkXNY1{RgtT+ zN>paqSyUOd>S2GWEPKV>Hk(XDqVDOqTAFT~DwAQiTZsup)g2$$xk34=i}lqQ&=E1X ztA$M~G~J20IhT_2*^MU>k{W(ZDR+ZqbtbL7F8_KruJy-ZsXLnqQDyAeCMOaXszoR1 zuCFy_W@{$WX(#%_L`5kU%H<)qYS(%*bvBwe(*AWQ7jKMfn{F0FxE|+-uY{}j{i9Js z$q#$+Rn9v}PHVGztB^~r*|=C7CMVf-HZmhRR3hLn^8SAV;h?hH(7X=RV7dwZC8pfTHBTiH8w9N z`|Hs+Xtc^UBID&Kn{Wr#qwSW3DYtEvD!IPfga1j%R(-0hirKpC-;KC-E28#7+^*Jc z=WChiXdVpr1AJa%LY-x$zHNoo(O@O}He9#F2m!xa+fD<;%1~>z;2+P-DjQ*u)JhSb zS1ZhS8-I7`%C+mNnr|d`IWZOYHb%8@zR6DJxh|iO!%V$hVA9-ZDM!oGm4B2BtvV{> zcGpU4U%SK1nuEA|v`Ee7jae*^t4}(HBM|)q6N>kF;X*L zpSbE(SMK>dkS-_0gwj%mOJ5@9pERQ0{3?-5M&hAiFYW3_W|dgJnRof~?qb#(n$$aU z@31e9BAHRAm({dFPi44%Ww&()I)1oFxm6<;90!u4X(Sjfl`7?^D;jn6xpc19=Z0|@ zaM7~7RKm^pqP8A3GdXXU z5tnbbolaNTc`(2)eUVD8CdzWZ7>LfG{ltr_WTLT{gi39;;az36vvDmns6`~DlwK7k z6a1m@04bXsR1(uLT>2p)v76^Ag>*V11_u$*%Yr6byf3z{L~GT6C&Vmrv@bulST4C@ z`$Nsw+f*u{WlL+j+?+Bl`iGlFCEMi!TC=-Zq&v-7#TCou{RJ0`4UzFpA?3MF%g=Zg zSd~$DxAVb*3#^Bt*DW`KaxxJcSJ_~0$MrqM;lNW@mHBAX%X$1-r@USDHc_cxC-e}| zD&b}`>UM2b>y_KBpk8An7%yqTVkj_ZhkW62vBDI~874X$#vIjyg{My{-9bSbCw+Z?p*yKY*E6NL z4ZD~1`I<>)Hn|lyOY^ySFY3+s+x0e=Z1VALcN6m^w{y=dSYFPug1eDw^hP`XDm)ai zDywb05-rXBA+<9oR%`1*ZSLz9+-kAJrz4$4Grw5Yv4V^8Y9ag61vQ{7tGne&b`NT< z+(J_;0vsE@VX%}Jf*HRbv%}SA(C{mLO=|ipVj?q3x4G&v>P=TxfdJw&g<)FSC1ZtA zEt<#%gmvAQ8~XfGAzYSr%cMK$?X=Scm$K_9opM7>ha#S)KcAk;9#?QStdu4tzoKa6 zDc55oK6X&5)W>NV(Ij~qPiccdl<6;e-i}9W4~NlRsnyB%xmeAAy(2OlKYu;ET+733R{z|MYxcgkaHLRx-J$1E-k6fvEw}had-`~uqM%|9TUUTh) zU~w7?4Dy>?W1^lXk_K(&BFR)(94d>=DpQCC z8`&b?+9dj3tvuM&%O$Ao@W}b6ZN)$8$HqZVY~*RReDzgy##N@Pk(d{tYFGV2UCY6#!#-I-F)x?qw>YZCTE*EXvWW$WF9{(B$0S_DrseVQp4g z)Zl+Fg_99iInxd!RHucRm3uIV`YZVs{FNg%Uy69@`EffQo+z8n!X@Mi>EObf33!CG zd%%^sMq{UV6=BJ&*+8JT=yZDUK~EyP+|paB6he{yCRNWS$Du?$U8;FE(PDV%W4dZP z*m5u3uI9?e`7)VxXfPBOK2I#-_Ak9~#%IxPI?#410%Dbg**VRZ*N6qkXnyQVW z{ly3>fgm>L^T9aOES6(I=tjG&FY1ZAJl=RuivSr-F|<#=N=(0 zxGsR9aY6jUPpWFO*`0^v{(_qd!_q7}+4cLeZK^7?lL2O^6bf*~LmzA;nq^_GM14X3 zD!Low<{_aD&QaTVLu@6ORI0N^Ky>qbXOV0#q-m&FG*qecL#mp7Wtbh24)5C{f=)@)S~XN{%LTXsj} z30{xtYP;K6<$|7|B)f*IRVPxJG`cB&ZJJ)C%2{uH!}|RBe5PIoJyu)ZO{^moOB}b| zptaxM8$^P&#x^)#a1@;nNB#L)%q(3kS2&bP_UenmkeRv?0Xdnk z=fV-sGTm!&{lRd9I9FuR8Le{jXm~1aR=KS!FpjO=D`{LU#w#PPK8i&1!&a%$4swzY z?gwvt-Uui|F6Yg{`&J0Vy-bYlxHfVM9$QxK&jdM`Rpi99vha>$5pJOblTmkF0Mmj+ zJ>cr&>QM9~YRj}Ft+(1}#07hSyxSR%gu6o-Ykv1QP(`yn9t*#PP5M;!W8t(R=wOj zn2xKPTvHTXUUy~J8iZW@)F*K})x(1eKId@N!f+_ zt3o^;ndOoZq{3lR}9 z6u*h6$!RXZZ9G+YgrGxpYZ?&5Rx7%7_>z7tJejDHsR+w?_PJXn?q^Eq- z9oa}*F&de<-Bk$}IQa6#5YH`IX<;>OW@1zSrpbFHe{>X_a>+%mhfuW050C?qLKtwS zRSH}`(3X0|=&F)RyT>_ow_7Jx6`#)+?sqzBqPZB(QoekzJnW2m@bIKkldO_o1XG@J zZXgu>{&qx4uhOw$q0cQ62@iZNvjme3kkB@RNQ;)0gvw$#z`ke%Z0eec;l)H-c39swVQ0jU1{aDz8IP9cH7!&74$~}o9M9O zVd_E~`2?J}OKKbw8%8s`Ub|7+CVk2vwvArHoj(M$cedCA)EdG*?X5*+0pGh%!(w(Iu`6W2@t?if+I!3 z?Y``8zKf(FS>mHuRjewKRjeXe=4@$-MKwf^<5^?hu1 z!q@}D?1%P@+{WnKM1cu}*~a%5%$H1W*C$hpNPF9J-hWadA*fU-rm~7=PDP~@CO6@MJ zku*uj&X}r~*}R>C6&giLnZ%*yzH?_|F(ueo!Jw?0JTD>(ovPysKlAtmab~47Yh;(I3m}?C*eUD$jHe6XCv``{PPutK_<2mV6VLWD83#s%m!ezr`V2Epsum@W$ zqVw^{S)Qg2;;jeHd0TC4q+pQ&9C7oxJKYlOqT6mxZtSb1VHc=w6=|Zcmgng(fuNn` z%pG|aD%KVf5L|%PX|`)clu5TS;soM}b7{r$#4XIt2s`K?n-A!E+Dau+Sx%r0*dZAo zCzh1sbyw~#;AKFf&|9n)Cye8|^9ZzCI#}Vkkh&{4-o?i)E{xCqSgPHg#Plg#Ecr0OFBwVj-L~o)(h@E) zi?Z9ZvpXy6b15-h7PLZ$8Ye%?#`ftL?sR>}XT6>@n6L-V;yZ@pl8WMJeg->y-;C~ zVyB_*ANe&y95-w zcB7M`PK9O^dEi;)LRb`55R|z^FLLB?S(@0)AaFS127i+}m`VM8hLcHnmz>H5k= z%djpMG~J3>)^))mj%lH;%Pbz@BiQL2IkG7X(Bn4v1FA*PF>vmEkaLiwHOh$M>EyJe zxo~}I?n}#L#Vs8c0||`RHX-dVdP|^oz)cx&+8jnChaQn#iIUvhRL8z-EM^lsb4t)> z@;-C-xf#g}Lu2JqW$;-gQ52)eXkjRe8ThQ$rwi@p&aonoEJ^xG3BCcw8_w)a8bX*z zLB`-t^L8C)gX@z^rBG@b*0R?D6ONg92XQHOO-eJ>xVZG05$A*Qa@i$A>HvYE80ggv zs}bN9VY|ibNU8yLCLQ&>ZpLIihvA=qPYvVJ)NA&wSjpySK^O*r&-ZcRe00@_{Zt|JppgTu9TVAS^MAP&GJ*iW;S zp`n~XwyhB=j9<8DQZXIQc8bk-LxvEYgcQgl=S^%>P zw?bKD51}v8)3tfwy3>9MCujl#`y{lSg=ch1$@3~7t@ORIMmX82%hFvA(}q@5e?9Ra zT9$K?rB<1<;6NQ}ZxbixNIlAfeN{t9REtEIvEXvDt5SNrUW?L1_N`;)E~=v`s0Cg) z?72UqA(u=5aR!(5P#H~^5T4R$TsB8jC2b)_w#t>znxX`f&B!5Vo7>B|=Fb(Eu8*(` zj)_y6ngJ&d5T9p9Ggj2#SkrDKows-O+4tvJ+|psGS5pY(Jk@7#f0mp5WHD8Z<}zGc z>P9^CTR2NfxJIFdL;i$@)HyrjI3BbRaarr9IZNia1scg3(j#NBV{wT2&3Fldb|Gvs zg!rOHFDG{C0zt}LSn((_me}odb|kp*CDJ2)E}$EFD8#0K>=}{U(?Yf@b(i0944j^M z;7WA7U>E5OM6;HMQumf;DVHtD3{pYTtfZ{ds>pF~Ky?#An@;ufWtCPks_AZy|V!#R`MejHyj>2}bhn?h+FNPh&f?f(rpXVaNI}0gYjv)?XIXo?p zow{h3eXQ<*7Y6bOld&0_9CWuWbl3%{OX;w4RJN6sU@Indan3cdMeG#fagtzn+~%W+ zF+bkLT+5dN?1*e@g!5d1ZLtw=(A?O@8*GcoKIGL8zm`KFkt- zzdfZx;~K zxMD{ZRV{Z@v$}N30nXSU1Ati@n`%>mduFrd(PN!hjlg>;IdnFcsb(Jpeq6{k!m67g zF&u1)b?f=dpmNy(#1yCOAmkSfUeKchw>#JqA(A&^TO$YT2zJiYSeZhAUn3SHHaTxf zOS9AEpqWyR!htJuE3fC6zd$E8)S(b6lbw|m1mNN+xA!`P(YD$0vK^cib9G=mN#nga zS(uH;3bsJ2SbkYM+ic%mCMzt4;Pl4lA;L9=Soqm^-WHVLpNQKAJzm zQR-0MofjviWh-76VqTPJc_JYmUmGWLEVpPrlo#>TvbS*JC3G+uba>7cVr(b+YHY*q zP2mNySy;kpwvl+yX@Yl9l}WG~56gBYU(nX@Czni=2ApWJ7m^;CmlEnrjHiQ>GfwZv z7gW@d*Lh0Zju(_Is#`ftPH4xEo63?fOday-NoH$*G25hxvpLBJ1(pbOqdR3A!Q(lP zv#6~-4=1G*8>4*FvdDzsC(sAt>Nq12nHK%MK`tk4(auaU>Sh;8_8>;KUWn0<<#9z& zZKs9Yk^wbhC9JEOb48(hb|+I*Q#3&|EBKMXV0*s-VZ-u?-c&Y3ctg;Q#rx3+#11&E zv|V68oCd^3LBRC)iRn8n7|B96nZPg*^|Hk;CoE0ue09u*JKBXqt33^Ij0A5)^Oedd zE`}ACeX@eR#}bVW^^wY|Ra&X@c)jP&qut~L!C5|U@d^0Zr~GMx^>HxzBCeQt;W7oS zTNj1pFFR!lCUvr$LEw$ZqmA6CUSd<$U=^4edtOdYGjzSY=9r8i#tBkSP8c6StkX5- zYO+r=h&_S?lu3Vlzn?lj#6>{@N_6)uhq8ik7L)bbPuWwScEj_=+SHa)THH(~zw_p& zkdYMsXOK4qsYo~fbs$SA@9DboAphxC|2$91wAZqhE02b5$cuoOg#o4Dl z8Vm4L^x0qqpv>G+t)<88j>7uk$u~|<^>9aj+UJS^9E)>+*eJlH1SH8T3k^VPdO)lS z44OR^W1k1?93lX7DR=!hAb%Qiep4720GI@BwW7zJ>8=XV!z%ygpg@%1&Jag9KDIIzki59VFF7cU%(J8l8@NfEoSlty^6gUjenI=j=$G ztOfoF0LX!~x_c~={yG$U#eIBYe!Thg6ZvP)SkY5aR|o$Mprb-+GSFW**L1!I77VCk zk@f(ffUgq)(&n{CAnPuRfUN^id>*>KN2>`TOb0Fj=O5%Z13kzcN5u6Fxcw5N1yo>_ z13>EW?R}BgV(w{4u7nij@ro1Bb4(0B%2?lI`hTPXVdDR9_six1w*qSWHO8J7o)JI7 zyt`@7M=9<*x8@-m7TOFh`|q~~ZW{6Z%Xm%AfA^J(@TLK;F6uSy7=9FCaMPNPRN!mc zB)ncU-8AT9l=hm|{7WxL$(!oEc>9I#Jx9iQAFI~C_?JL8|J?59A7s;W#eQoc+#vTX z|H*xx`h(QE!d+qCRhHj0?7P4FLaW~|S8pWIFFl}0zrCr#&BYhWe%_<+!y5Yhh|1>dS}e8&?%1t_Fypqu)u!-K)jHw-`j(~_|?C9#R_mJy=kq4&8Bg=3B zfZbqz^qgDLoxKF+2$)m{fH~-~ya5y7cL9gb4+0--ruGd0qE}vTwe^NbegpzyFi!4_ zlAaUyN*4^z-CqLKgu`>@Er9Ural{|>+o6BdDX-jL_ZHm^niD*l!pGpD9wGi}g!L8f zYjyp?1IYLrKve(Y?fBUA2J2^V3z`UE2Uq02AKcd_XoDa0O+UIdz|KICy=_;3X<6I} zlJ-h43)a`+2+hl@3<|s@W-v6kC3NmT127O? zRrD0a4m<#MC4d(m!8`=J_yp|&i?}!4`a*yBL+@CWH<~n%U8kn+pn2iH|zIx<-E`zFA2Pdnm-71c*gn} z&F`D0{mee`om+h)=KtEww&r`RznkX0wxrjzLVJ%G_@lq^?>`BGKl&U0!H0-~n`TYl zQ}lP0Fg>{23v=&l8=a?TZboYSvBc9&<33)QuW1?iJ;dcrBfi(`FVy!rEFd^iit+(r z@9vi;`Fnr!SK|NhJIT;HD8O%&p>H0s`oFko-tE;FYW=N-&F$x2v=L@tBY+d5ha-AN zM|t3#3G%(uE6A$??DapG6QHlI93aJ01Pp_*&add zUe*1xMfjjqzj({{EUTICh8U45{ zzqWF3n)NI1S>P}A>NCsvt3Q7y0v^Az;0lZPe}AWD-2tqAW3xVNNU6uRe;1MT zSnaRA-}^62gKt+nH{SD_C()9WG26-U2+3KV7DzWEL4vER5m zqe$UkqLHQapdv$t^P?lkBP9KZD2>+4%^BGhmE3s95~ z&VZglyXOkHJL|a;sP#4Ib?oB`1g0?HJ1^j=Lr@eSxV8ON=;vF%+w%nqlcZX4C)T z&EmOf+_TFsRR1OW>uEf1w7U2a7UWI6K1y+~Y0qDJM>sqV%~$W?m?tIlU@C4ciwuvMD35nuLs>A1FxRTd;i{5`hht1jwv`)Ayn1x zOWMQM1-zI5-`eBhKkDs!zvI1!#vZ~d?!ym&I`x50^y&)l_w=9Kg>&?i z4>s+9@(xq9-)Og=UMu>xpTcj!`ncP3Kh+2VqhUSWhff6glhr`Zzx#5%5qRJ0br3z> zJ4)guf&IcWQonv*D=2QN|8dIu&h5ETK)?Jr#>F%Kis#o~_`-7y{XmFzubm%$eH8y! zWBcEIr_(>ka{D_^mfORCTrIcv5T`sm_h4N;uj&Udr2T`BS_B=fQMfg zi}%moo6#0Of`+-N?MEr@HSPJmCt3dewB1c(USHO0TJl@pU+vzs;WeearuM)0TFrmc zhS!(%nw9|m_FYKh>rIfGHhs)eU(>qZeWg`&)1+4y^_q74+P7L~H#Pl$hhM16^CP6c z@L=mb?75#VJnFcwo~}W8|Ic^QA@|2g-=ysRm2WUGZ(8|{kG#;f57rlx4>w@pIQD%e z>`hHaA71Ett>WJC`>v}0;FeiekN$_7)NoJ$Eqr$R{Gwrf)6T_r{KU6g?XPO_2Ac5P z=7$D-6s!NH&Fi1Zw&GRo{>AH-&ke@@pf$%6yz@oVsW?*WkoGC;ikK)7sx80C8Z*?IG0yybmx z3v6WX{0jXRx%cdVzvI3>*KO!y;rA!H&8~gh^x^&egHQ&1@cI!r=>-$LC<6!1z3&3y zp1=CLpTds$D){@O;D5O4iQxXUbj8zj{nJ))n99?47S~JC{e@?i&!*PL8U3f__p>Pd zg|>eCvi#URgL(0~SX<@q{hN;mLQkFd zbZTb@cQWtoLsZB!=>~GEfc&(pBh&$>;x+RrDEHo7BuNGE2H@lM-}KR|(CbGCXYJi( z{j_x4;&=eEuFvm|>b|+yH`nBm?DL1h-aZ!g_7iX@^ksn~@bgE1J9t%l$MW5moHtPb z{dgFx4%mSUv9AJ{a*kg^C=Jk>!G&H9mv0|# z0*SZ}BE5Zcf4X@DpCzP(`MulmAvNR46Zm^8DDO4heIx2#Uly#$;sd153(x%a<3N3n zQNL-!>&p4g^}q2F{mwT}nGcwHH%)kDL9c4Vs}{g*8(vq?YuccG++}#vsE=LpYg+lM z@0AXZg8!@cKlefnzFmX<*Z=UnO5iIMPQ5)4@ZjCrhwaOvU{|FQ(82-mUZW3S`NPzp~o*#Y$Cg2D9-WTY8pK<+u){_#0 zhX3T@h0ohh(d+L(x&7<;@t*nj8ObjY{l3M%pg)A>_kFh9y9~g~hn>~c{C$!4_s>S^ zPy9$duI#JX`%k{b?CqI* z|M0#XNT>VTH*5Y*G{k@B@lyPE;lIEC(?^2b_&eBwP6y7AwK2LUzW zAM|CoKReP~9KT`BeZZ)Q|MO?Hd=**lAHD_2_>E+_1wdsykmZUtz?oY;c*Y+SS+0#F zPh`0O!R`4Z8M2{2E>dg3tocG*oRCCl_KV$e!UNn6;QjA62ep{zqb+|X0k~--Xqg=! zA{oeYM-I*pjfptI#_r8Po@iXyKlyum=B^(KoY<(;YlYF(pBuD zz)r}?gW4@K4#}&^u4?j(05t2VnmAMAE$@##c4Vq8l6)N$ivfkIy z;xt>gxwuf^$4msUdmv*p%t;QM#bY3E30L~GT>|al`ZTKn9F4)proys=1{i`?w6aM^ zg0o1O)7g?rfc*35d$c#2B4+|fd($bNEm?@J=ja z*M)k$U@0-&0Y;78rC3TR(!sLOmxbmKA_%$`dZY-@~Jw4%`K!j%V1JF65LI2t#V-3Fqo0HzN$Q zI>7`mPa#;bb9C4mxlOLJi1Ub5hve04-`<^FB2c^@w8Y zaaO6G#0X_x3jp+_>r}$dCd-z@2v$u6yaU)v@-kz|(^%E!Oth|?3l8utrMnUbV^ru` zps{6E(G<}FmC*S;T8?m5B6+oNi_I(mLKAl2NdhH~xgp7jwOG`(obfrp=m_-Z~ZPDYX^#B z^@ap89fxxiU`RllNg$LQfGwlrq(gx+6}Z7dB?FS2$>5-jGhnE+U|4n~@S)~JOciZ7 z2PSTnBO`>>i!@b)`ZQjmTP;|(CuP|83uU_N5E0eErQs^!ve8Bwl5u#*oG%$-g^1oL znouE$FYCw(2rGf9ikc!*Dyh5io_2sBG+H30Y?Is?M@Y_f?jXh6hU z;byt0DGtjtJ{ZIRtA-!gb}&nT^Hg>V*>8e^76CWdqVw|^L|7G?q%Z?0`mKDLCML*w zWty5BLsp#rhoU&6rDL3AXS6en6VM~<46FaKH>_P^CFiBJJFb)S*4$0jhwG0&QDvFy zY_lCFph`)_PhDD@T$t@(byAdq8B8j}SCZ|HvJ8JC0@z>1bI|X!T8y|^Wogo(8TqNb zodXxu$p@MdPpsPVe5&#RApgs@R3(7E&ohmT42NTNof@Z7kS(ABB2|j4LFaR+s$@~? zSu!xlxk01=Ry(m(VUaVybP*tH7?<_DL?WqHMwL`H)7l{pt{IL1yRts8@Wvk^E*7&cw6ZvpsK>Zd;QB7-z}H000|T zfD}jteAJfN4p%P5R7Ybw0e+Y@+K9N4X2_q;dGd|(_h2f-Xh~0-L~ZSDimdh%z}8Ix z0e6K?X<$pwT6a?oP7^@ucGJyPPG&a0^(3c}Ry))Xc7Voi>gVY(<5N~9V=- za3RWiGvZ9ryqF-k@@T9xBYb6&XrPWGd@YbbptlFyXFjOC)ujnIl!5^|MHFG6Qr;LT z@3Z=HCt{}-DCLD_m}Hr|N_OkBu{%{R&s|I(ct-TamIAQh!O+49Q?-e6c}5&xOWmki zpXTaDem1X`%RHI}x>t$HFk20bNs&hup~S#mumU+V66}zgCl;6w;U%NI$j4#)Q(z8> z<%5aa^mG=fBo8dqoN1dy*+4#GRlJ0tUR#GNVAVP z#|SfflBCX>(?-k1P+x@#MXAH?JPZtV5~R|mKzV-3iu%4h8Qd7y7xhz4YSfH%W`15N6HFgCTvsUvZp;uIPhzuT0GPMf=oAq% zx8ZeQ$8!fKd1eFWnzN+U03Qdk2h`Y7AjB=sY1?T6NcSpXN}P<0%L^cmC|cUgJ;J&e z8%;=H^CGTJg)~-|c4~vn5RnT@#;`xgMIf|3beAo@J{S(50h=c5e@fImSOz{iJH-I( z-x(nI~sh4SyAWKU{h_0P!)Y3IIlS;?MjWUq;BhCQ-B?nC7awZ^Mh5)_rRl+ z58@TYde|^-Or$y=5e*>VL8f+_?L;Dno}i2qW+9uLjPC2A1^X>thDwyq;CjN&XX!)& z3|GS6B*5_5!Iq@crZ`~33TC6AP?6bcj81c7c@nHb-4F7?91$=4@SyQvf7IjE$Oqj( z7rXTqQ%&3ngcXY%Wsv=uc1nDFY83R$JB*ff68S!^gvkL2Fj8$bPy`MveQc^&sn$tN zk(?WO7U_U%yAkL%+^VHi6#Y)8270>*B9!dlPS?-!O%X>XA(y6#hbABJXG-l5;b7x? z=R7!FQcj5XKskF%jHP}CouSe2dI#HF)mKM>ku{hl1J864wcDPk1i--=WC)P>2GAtt zYeHPrg0zwHOG)Ue3GANd@(`LFPzk7+>1N`>znmRT<(9CjMjj#g$RgHDemnIXLmjn2~#qN+(km^LRrNjlm zj5j_7`{m9E-?!&B@cFq(MEG$di>9m!HtlL!C_o~*niNpK@&oqo|38)WlC2$thYIFc z+1m8%;N##|NXu^Cv@-?dbxVyd(TD_J25=5_7A4t{;%PxS==FdG^ZSDQy6Y!{}&?N{4^L{W4PY}er1aFewU zGVMGD-B3Gc!sX7G3k@4SWIY)X_8Jq!bk-YK>=H#T3wo(CQM&4`JX$Vf2BEeL6Visr=?eKA%{q^#FfR@7U%C~6viPB7sD`hDIq+bBT%U5%nRFG33<^ysEHOZs(`xI1O=07^!YoS+ zBY`{XgYr&TRY2Lsq+B>1whT=KZeWfW~jS&pLt2%2eHWVu8|>pBuJW}~aAG0st@ zjCYNNE7oyk2Hj@OO=g0p0ADm1#AORQc3qF2&E-i^1z`caNcDOUGrYb=SWq;H>^0?n zKvT#CwhHuahb_%oZxx`Dnis(CSuR_n+)TEz$#mR)x-h#LzuE4#8IW2r!*!CUBVUrX zAe+~95GE@_tSrzIl5WBNL+FNua3-vRc2aT+80Q|+KtapsC zJrntbE?}>NFqyaU208#!bR`ENoc@@mhj86H5wn z!Y@-~Gn;iS6&+HTsvhEKp(YI2xlB1SxAOx{#q$_F&O%lju_m(cyu#xU6C?&(jsUUpM`XbqP_o@IjUU!1*Ud+lO@X*k(B*MF(POGnS_^nifG@UAM}|#{bJ~Ihjfd_BgEJOtBYmZ^Z5d0 z$I9gbv;`^AM5M{$CyN&{e&q=PHJa_Z3o7fPK-Zbh%MeSGCxyU7`qlvk33ar}y?B%Z zfnvVov(duMv~V`z_UPoS;#)W14s#E-phqPyLV{t(EZ%y6b3Kho!gXiB7=kc3qRhN? zi*TrYT(}OU`8iggn@R^J#l_eLf;H49uuDbN^}Nl-osSUAd|1}fSV|CiY})e;5RN&NScAWKoC9fuTt)gT{0RM^O8}^pFToP}{XWC)|K5i6IA=8rP$hlF~63DZXn;k<0 zzjPGJpcjgSUSF%zOMTd%c^ROV5%x0IsD`QY?Hangy5O=aD;k0lnUmu{*ctf1;DL_E zHVQ)UBIyegXfeAuJJqek?pDWXp3d8g#*;1?yQ~#xIG=LBf#H$ygyNZKa{=)2PTZ6L z5WioB(I^22otiBev$R~pr{nx&cg_l^y>M5oa@S{MS10BoJ0&*3kc(r8l~O_80wsxj z(CuPI(YUsrBWzRR+pR3K^JI&RRqqsoQFj8SBVbKGZ#aq^ISgV@^uR4HtOahD)7dOT zh6i0%d0@bYR=ce)awBW!!Fh_P2kuBzm;92WMmYmpR!(ewj{B$`vFyY;|_xGMekzBBsO?A^bBV=WD}3G#nR=gLv7+B2^6yWuF}Rc8<^Rb^{$$% z367l#IXm=yc1j9drJpmc$SQ~OEL)j8=xS`#tK&yJkQZ*L>)p4b>M5UV=qHjmqi0ZOgvu%=169_^g+X&dnerl z+@}5hw-aqu2y5V>fBJhAmy8^%p^OHjDwtNDW0i!YyvNn&ntG)c($C4e# z#n@9zHIaccnnvX=nh`(;$9iB}Yc669z+^qv?b^nq6i6t2dS$MFvv0c!E(TaEI;#Oc zkYZ;!GiMQZqYjzv!NuBC4acM9)8CnKtWPn`Td+Sa@{GX`lSf}$HV<#Kxo9ppT%Nm~O;L@@(DPZXU1 zcOpw>Q~@-V-PYgWehM35&zu@7IKOfAIY4f0TU496dAX5-tyRHot5Kn_aeN0va z?GHxN3W$==o1sIBhg4~(CC9E2phO^Ju|cysjh*qLWWoC}B4&&SzVOwSnB=USMG9SC zmhFWky?P^e6)%E(0sG#CiCv9CIE0}1jEDKlhHV#3P@_U)^6Uv2s8^c73kyg>(HQNOd!8std;TQ^$8}Z(^H?l9b zc}y~f!dc;9hHF29Gqf$Jdq1^|dJlZvz`iEIwgyjSE_2T3RR@lIAlU388@$EmieC-_ zS{}!TIIB1IkOjVz5Sh^_QS!vqc9nwWxYf^9Y$`Si)ZH$Dj%6ptbNZ;BOC6XRbzeSg zF0D(iq{a$Ht%1^YlJ1w8Z*Kjvbx(Va9oeT{GA5x{G-NQfuxbU=+m@M<2qz>--Yf_> z>#gD-aa7og2wH2^%L0n8+%dxUsvxi`rjzWu%k>r z9w0#E6)=TjMr01>8nQiEY;{anXPN-#TA8exI?^7?^8n{O>%xsF4Dphr>=DMo@gmD>HUBBN{^GPKOZuStwtpjMgXPdp6Kqsl;GZ`IKt98e@9<>70y%Xir3J8RH zgKiZQIeG`sEILCW!`Qq_CU~89^;jV2f>SS1J zT0~EzJv(qP_ko|?p0t?hW2$1bD`iLl_s+tbi)W139W6{yPT=pZL~@TTLW?y~cE6Ag z+GqzX(BQ)uqtTLrvG;*bEVEUlsby0!*fPGJL8SEdgU=m&8x`&;ophJQa6Imc}@%0GRKMtLWi<$s0` zel42io|WSvgyo4t<9V21{J*^W^7rTD`0AfY!3koZVb=Fh8GT^jW43*t6#hsI(cj6A zUjthHtWQ({)SWvFV&|TF0?;)8%prp0mFH205Uzd49`P=51)xyeCD%aCO%8YyeRZxh z8UTlS2lP~wmEXSo1acftyC*n@9`@o;J%B?4YV>`s&NZU!DQ;W-IE}oCPW@OVLL}rb z{FUeFSubUfyztClyTW09?;gKt%WF&e&TV*z@Pd{OKgz)0H0C1}_?kBT<#(syV~E^W zU%U0fcb+fy7qH`%|MmZU40n4mG34hVzLf{~l^%r$5-8s0wHm!Hr2 z1ag3c>9ZgpcKC@)<1x7p@Yc%qi#PYd#eEjW1940MjIaRmaupTV#dk&jc@azq;*I+( zw>`Xk1uVVKItT#21q$pX$|Iy%ADaA#9_a@)@bW;_uMx|Ait3vqJe~KPHh{xZ_J~6u zn9^$z@Kz7odd2Lyii`&?pr_^kFTd0oH*@CY`!Fr)ahfD_?lKx-%Hv&t@ZEs{tGq!c2#gQCtp$pU%{Ra5`n=+ z>sifU_}@H;@%t>|2bj34yzf(s-_;RMQuHssQ`F6%eeD*sg3 zS4zEQ0H}9G57PS1D|bnukhyyQAwWL=@;>SE%c~*r+1s4>vhaZ5*L+SF?x8S z$4me6A(Qp);+r=Y$c}lQYx4-xrTJzx?#g{X60@2m#9QiWdkGzs}p|Z+LJU9{$$> zEbk6B{xM_q&I)_`-aq#r`Sm&U-}4uL>+NrTL4?2l_BWsR`L)nbT0jWA$2|D7|Mn2z zeqkR6w&$aXcO9_bdW-ik2w=s9fEU?A9Qb=c41_uGZGRqGZ$c`w(}OVFm)BS72T-Yp zDe<+cJZlNekb7hF+rIY=_aCb9a;bNH)K}*^{~l)TctGPKa)~m(;d?y^ zbSR8&3zF}C4ZfH-_-9h;&!p6!NvYuQN&bW1_+C=#aZ_-Y9SQd0`#?dyIXm*v(0tf@ zJvTMqn;rQ_zy8lXkf{CPeK{YP9r?*)^5b24$zy?)QKH%`@$KwJ?4QGLQ-K z&AZwk`+{YkKmDlT`!_KB?&gEGyHl|J9g=%?R`5Vg8u%K$H3=4>VvF{&t(BtQIQd zdQs8)%`Oq%#kA`qeR{khXpbpEP|=n?I-h@K*q$p8OrKtPBD@Ji99P& z7%q01>I~h5ze_gN=5RVH(qde#7Rn{ss*7YYnHZzRNF30$u8&Q+>N<`#0|-L1w?Xa4 z1X_mfRCCSTB#Ep`QjT-UFt#?M;c|4!53(#{!f-h`A9}+2bvC>b*2`DIdd2V-$DZ8Y zypr1=l#3V{#^^K&Cpkm|j%OHuf)He7;5GI*G(rVQ;&L+E>okcSf&;pxPjY+o7As@A zRdU{h8pmv1?JrA1ri2Zj@@gX$yvsYyXwpII(8SwWDy^;gII^3mRZ|_twdxT9p)b39 zn-$ZkLamq5TAV^m{HC1+RbkJ`h>oU5-sDG&N~uj^IWD>+G2V>55$~kLyvBq*;!iS& zp0k$B9~dh|W_Q7&Rz%zGqjM9gHAslMG2zVCguKpEYXy=#-=32ixiC2KCt<_2n^Lj%WR&Pl z2B*@B9_=ETk|aZbh=;?#Ah3Nku@8i8XSy^VjKRMeHZo5O(P8`@ zrPUW&XQd*8lu(~ECWE?0Hi=BD>^$7j`Yya6OtNgo3135-6*kc!cR{3wV$-6eP;-^U zVO-&x!~ybaIAfD0wADb;n?Q{{QQ~1_s)?Hv3pkk71&Ar1^K_(9S+H1c>&++{c|yp? zyx2&{km7~p7`TK3nH_UaLc|ddsXQBE04ck3Pe8<$&lf@<#cn9VNj*RDq!hAGrAX3w zH-fHYc|_dv+CoAiU34z9OwiLx>^fG4crT$$=R7&A+v!SmA@qO1kKkL2c|taj@8AtD za-?tgk>Id8rttG=fGge*q74+XMo?a)ByXHGXfoubh;N3haNKt%*on z@Gc|;x0I5z>}X)G2cpqn{#jivN@q8k$Qd*0RE8{Cf)977CwUz&M-vE4>>y+h(yR0B z%9dzUnDfp2Qc0@;1Uy|TJ``$xGRQTAfm|ks+~)4MlLs?R54R97Cj~rN(KUUVbEU+> z;_9MRBdl~LqEk-PB$uBLPT_*#ZMxnLCf#A{#>VqeZ9x?GLop#p@|kHe#xU&^NfJc z7NWMFMd>8L)}pw~XWWvj7d217d_x+(kjlxPpym)Z-S_8)Pfkc_T*^`)dCZ;;P&!XG zG&)$OOKf5iB_$S;S7n+mxZ;VaV;i!Ff_ls~4LX*(gN5b`lxG(hYILZVbVokqj}<#@ zV#KQk5FB9g3MF_vs*v=ERkaxedc-^sK-73+yH36#^!eDfiC|ft7EW@WumT~TQUfo; ziJFWl34#S9vl*&sc?uJ%5jG`W7ncP3g36LbBuJ1SgJ`1Uc!aqS=w-xrGz?v+@cbBrZ4Dzaj`Y@6A$AN8P?|{s zEY&95L1!brBOv|PTtH&$3C%a{LO)@(S2s%JA_U}Rusnn@DUJb27`Z7+lU0Eau(Amg zy+x!A6%S=T$->)=FJ!z<@@1wvC#8*#Ou(OIgZAI4Szb+ zQO*Olz=~R)^$R;vXJoW(MtFnRIWIT|h>&X(N*Ylaxz@=EHeg2K79!uUN^(XtGB!4l z+dh|=4Q)2OgQbXA31T%{EcsdEHgk06)klyevD3-UFw-Wk$CKw(Gowt1rILWZq%AH4 zZ$?vUK8zryN+c?8p%Vu_T}Q4Wu~SMoc97{mn2E}EJ;C`CKUT`dO zcJ1C;99T@?+2cOqu(0GH*L+4SIr}uMhSR}}F3RTYj4~>mwUZb-;oM?Dda)YAx>6Yx zXF`a=+Kvl)i#giP=8NViPWCgT9T#4+Iy80wc7aZ?ax>#k!*oUOXqr}J`jC{vw(6G4 zfgYw^K_WSpl#nlhpD~FmZJZM@T&&`pZ?}3})84t8S+P)z$76MVOeY*t?c*7pZA*R2 zs6|n#%hjAiCu}6KZNU;Hcfv@Qrc`M`9@cwmIarJ!pSW_p0fa@SQVCXrX0GZGJDYq+ zqM)ZS#KbgBl6Zfjr^nR8XKb0UYiu40p_KcH%rB?Q*q}Xe*ajLYZ7;{N3d-8fk6?Y2 z4ZW!Zels3jklCP}$Cj%O&U7IL>#Zmar2WiT?1wuWsVjl@YJb>eC-gudy51?ZT$gSM zN-&aVhkQ{@5NT=|NZtshG&?7|1m zhkU^q8VK9qc#xGj=@=s~O$wOPD6i~|3v(DQCB>}pq_&n*pQVr$Rb_QIa7W2e7|F+K z;!n>g*_oo$bP3)$%gJIb4!5I(OCfkCy1WF@ zS{R};H8DbcLOT1?hy-0kqGv!IvFpk}NCH7uL&%~y&-c^`IqrSjS%KweTxeR5;tIlT z>k9>8%4m%ZwyC#@q5h$)$zfdjY?kecGE zaa0~~v|beaiEfS6;SjfD`-0>VCb`xm7?JDgIOyEnlH9F^olji`2b`LQy5{z#6}noj z!9L}-zEiMbdZd`;P8I11-v|sBiSd&3;UvIVk9L#gz%&dsYFllwE6D0{7Mk@48X_!) zM&i)~w%IJ?Xv(pdg~dzbLiE?u(_}eXD(XqqNm39TW><29`3T@AI>LAJXt|zT#HEvF zCfXrtu^6H#lji(tG2&PAI%>pd1mkE9v$&}>Xe5iT1Jumwi#uEG(Qs8aapqwer%Tx> zRr@)T1dv)1A}HzadGn}GT5S_SVt5!N!-%2}8+Aj&+B(N3gft&@dlRW?*D%#wEzg*^ zijT)BpbD&&c(&L!`gY-r&H*!82iJPYqBc8m7ve%R_zoX2$cA)Q_KqA3uqv4@w!%d4 zwy9!{sL_O;a%LVgCsj}`aelxY2xIge5)B8C^btABc!|ww{g4*%Y@B#Lz}dJdY1I`u zOVUY96gi8|^%bIrx{U*{!aQ1x$k+y~nW)7=0u`cKh*xw2AR_>2qaQkcxo>y650Sy6 z350JCG6%AW2(7Y&A%ZOHEjcQVQ$TYW8WFCDGkQQcRTj5e1!0;9WQ{IY4TTes8RFyC zdLWH9yRl+l++1>&_-ZzxMXlSekqZ^+G_uif$7IxP)w#HJjKpZO!*XsOEabFMIjNj` zdMe204uqCMxzkjQVD7A$`FQ4#AmC*?-^>whVzVtblbO>AR7NyZ#?_K1ut8OJ&f*g8 zU=9s@-N#k_K!E{MoEk}c8Sqp4RQkThOK}`4W0!3jUR@pLGA)DkVTnFK3_)f;1N)sg zSdxmd!I;FnOuQr5JIbcC$jR|=+$R&@zgmovT}tWe{6Nb7!p7AJVJ%zOlp~PVGNW^d zR~8XoOxys*P}0EmM2|vki`&a_zQxx5*l%on zwutzHMM2PY!olVnB?&*xI-@(Fb1J7m?Pz<1omtg!T3WH+13?OsSB*oXpQ*&W<`jIcj8cXoO*cFhqK; zCd=_@C2VHRaIqcptF$A~0Ihsfs=H>#VZO#9;RHIW>(dL)>UIH!6Br0k z4ve5<7y*I=83gbk86-fa0W!$g8O4Jj83f3{lK{ye8RnPv+57B$YFC}AuDabGxPXJ& zqR6GS7T5Zp>;FE5j|P^l7uO>a3e45_?9Bmmv;^o9ng%$1RfTf1vTB=Nw>fEwAlhAO31F1PEG zmfSY)q}!--_y{lF)?+$N_F$mPd|_A`s^0=t&D6-4C&Pi$9EtJfYV?+6}rj zp{dT$5PZ;9)^Xjcmq01H%|ZZWWt{igk)mgFk<{}m;Spp!*Wd^+gRhI@ev6Dj9krW< zvNK{K!O8O=YuBW3n{AEF+8w4&{Q>HLuE)u$xg1AwdO97%t-C2EnzZ5JAvVsv=&D_2 z*1|O|+iE^oD!yKNP|a_uP;%jGSTJ4LP*`WAAOqyg?3YS1hjL6UZVOx@X4h@NA)&7W zN=cPst|>}zD_5A+Q8_jYq4k*P`9kCM2-bR2GIV%4+-X&rB@uGO)e5AXxy${PwFgzS zR#sj@(M7-knVSp^t6YiRBw0Wj+qsoJxuodKuh8raF(+`(=*B#6;)MZ_9Z+v> z7xQ{U9h2nJDO1}?+^(;$nQyIy36iIq!%UPeEA1SGCmacLo)Os)xsF9+orEUWFBk7@ z#{xY@c{ox!-!1|o$%mlbn?p=)k6;!;9py5et*}hk=1s|OU_H%dPQN6)POWj;oA3_w z9W5>mAIW{s%%};dY9?JdO|xlt8dE?~*$!Vdr<<)M8tz^IkB(#&cA|^^a*Ay2b7^P? ztHq(N$QkR>n%{NGN-UJ5;)%uOLU3!m-A|+>YPeE3pc)F!c8S>MbkXk|p~}XZU9r28 zK-H$4cExD!9Qq{evtBQtbQp6t{c<;6mwR+~TyP{e z^>&^xxlWTPUg( z3m!`Q5kkU?Q#KFLF=j?R+*wBy?0XGyvg7rfAA6xorS9smx8Fz{%-@dB@5Q{!N!&o~ zbOd${vk-kGG1nWIX{z#8>v3Yc0NPg?Tn;hFm<`Zq0yRV@oFyTKC3#Ykp+~RV=b&av zf3@;ynshhp1j@}!b3e(d>1^~y09+ENKs(6CV+{xq%MIvAnY=(bVZxN5wpo>ElSwEI z)2v*fI0~C(B$7-3lwr;9j~YX=tMjEG&K`lF^|e1ABM7wcAw~-{OS4JSC%v}f$=oGs zcCR*ICY}<9^^@K4w*N3)(fe&7ZW0-g;@D_?$j@QD?;(h`CZQDiIyTHIqloKl&a#`E z9ooA)-k)3YByA)+au~)w?B?EaUg=^ZK-I=@9ZiUKun;-|z0wg92_<%Br=d(SC}N)z z6A+nE&-+_ZCTe-i0f=XQzyp99(+O@$BI;FNPdhs*-8|Hrg4V0$j!G7*dM=T2u!eH* zz1{&{fK`_926VZ*y7oQ2S*%ttq0{*aAO#b`BZnh7oA%OSNzqh;piH!_7FM(au$grR zMLjiK>pA--G8q1LUo^4$oX&oDU|va4N+RmiuiMsk~- zOs>r>6w7Ums>Q71j39@T0Bah4m+OuvL!)yuEXBo?StkWEK+KcfDd)t-nNINld&cE9 z%EcNz0&f}5zE2oi#RPJQ@G_}%8o-12XAiV9qN*gycj>iCVPn&hXNc{w5NqM_l zumBC#iqrUp^LVf@kDQvZSZKCb;EcI(i1_9oElcja#j+xHD3;{vIv23`I@dN%?Os}H zeZfWv(;yPuoi)Qj@kP0tR|T5Rxh~wWl5IyPFSM;G<+Ehe)Y~;H8&LeNw(DZWu>(Sf zA1r;dWaAOanj) z3Bu_z<{N(9#nDvXbKG!rDC95;tWpOFIN^>w4%h{GE8mP1~)S8*HLsQ`%*mx6S(GyDm zY6DmctVh;Hf#CYcDf1emTB%;a@~(-;6v<`8?Q=)^0I)e&4+d_ru;S#RW1 zPtZ$^<#*)*PvUSr+Xgz3(*W2e2gnBeXYR17XcPCz6kc2@psZ$`y`9!9F3GmjX8y9$ z)gDm;?2JY`Jv4y60VfwcsL3cA8xL8wD3t47C!V`Huv0+oQX&+J-F>8?=*gwq)$w_g zN)Twp%)PS{cZ$_|VlW{OlP$iDG`4pyOHGFS1XND@6W1lC99r+u2)v^n@-8_TT`TE8 z=d?gBHNeN(7#lO=DTY!UO$11xNS4n`*!NWK?=NW;wc4Tchb=_+xjoqJlZZ=K9ya%T z%@9$+*+b76?8F|@rk{{(peU(DgyJb;rh=+Q^3<8HB|U74aYu6#wqo;vCvDut_la{d z-L<}`O8Z!gY8Fijm*26VCdQn=cPmzGAja>rE03K3ydyOpsKSoB8H;nfqdv!M<)k-M zvRcT)8hXMH%HbZ@7Y}oYE<{I&Fr2M1I0oXG6|HtI4Kk}(UR&njnE_!iKdqBJsP|Qx z`DJyq_f!T&;71id;`0NmAFW|BTAcyN1FK_8c};8r=E5n=ao1W%>1nGp{#CqK@hn6X zbw9=|Q}sL0^c`M%=Dx>UZ5AH$st*NG6&G_Fai>}_A4`aPd6k^X%}&p^hIwLk>@siE zj92r?cCj!ka%WViBhsr4U(az8Gek+@@?{B0J(x5?-V$_I8A*^3g~z z1?I9B3juOJ!x7@xMSh4ns8biz?4(#7PA$wP?zNH(lWB=~1>p^=)m%2zh!!BuE0V$v z@;+ND%p4qG2eQo1=o=>4(Lx*AUY;gPV%AkO2fo5m?P$!yrZf#Q@S77ZN*G)Ly8Mz9A|XsKDQ3Aqq|R!c{!*Yj{q`23wsn1( z>;TKdSc}b=TdP!J(`kNcj4hBgUEJ$#FRi=fd3!3=1!Eq;nS+E*0Zw~|XqRZ!i>^yg zy96N)jkjAwtAbRg)!z4y*4coB@50dR787AQ4v-3D0`v&BzEnFafcZ=y4{e!{{8?Jc zA~RjF&bE2IJFDz!MM7nF=Ifr1apZK%X&ep}mcZORZSOi*yecJ^n~T`*hUMxo?>cfy z>ep)r!EK27rk6}!)s*H1GP+ln%w&p{LPfe!Ey?VpAC*#YxB{{aRHM}RMULx5Yj1`1 zkT6;jLzLpyrwbIwo@TqWU1|1);7vawjk#m7JMp+eNA0w`s>HHtt*yfukZtEMzye(@ z_RgV_8?+ouDYzVtI$+phg3lkSV0m7jFVmB_TO7;s9L|$$x7Q6qxRwM|+>e#s({W`< zAeI^w+Bb4Mqj7-i0MZ-Wv6P$wK%j@dS7qU@Wc3REfpdXUb(EL}h@DvUdF2#F3VoO! z7yF!9R)@L7Y)LVywUwMuTsJJB*n0!{mz5|(+_lJ%vdhna^&^MWB%4N_g^`D#=MF+O zix;NI!pQ=jCFi?o0Q(@E3a)(xu%y=Wyp5|N9mIz*3{YR{lesa)ec8B?I>bk&nn8eg zHe5JSz#;*6BS~FP0ai=`->QNs*c`9mnpNN%$L?YcSP6uHvFwgJd*v2PW0fpO&C$j_ z2K{R|1`6@aBAT6&TVBGCh}F*5<`9KZDsOW2#d7K#m!53QcN;%GM7v{hnN@g{ANYY7 zgVNHsCfV#Zg&6UwBJ%SKU{b=#1~q3Petb!o;IQeHRJN`U_nliSZS#)s=A?JTkh((j zWZST03yqM43>CnS*dnZh6vS8dz6o0K>*g-5d{TshW98y{$_!b&vEGgFhOmploV`pqP4xho4`SvBiuKfoa>b;b=Oei#nGpvIi%?W6FDy-*FUdJo9?H%lCBDb zVaRb_Ent~UY``~;s%i<*A)<*Bqi4lvp@R>B!oy&nO*X6baUk{(af=EZRQOCxG!mkW zb|snHIPnHP+MWR%7chw`WF#OzH!j63YH4##JE%c{?h7uw7%`c|kb8tIcoC^8hpYe; zd2dTPc+?7OvXJu0t`=dt&D1MkYUYVkt*|*WfjE*BuiAj0jJSHmv-a%4nE`eU$xqZ= z&J;el2K&PUc9(y)RuxA1N9Ekfy2zHB)&Lbx2%Gp^lwL4^va}DZJj}gw;8O&pWV4my zQ%a)iGX}F5b}V-hx|yD&D}AQ5oiT#@>w`b(0~{+$fI+c`CJe^N)v^WHTeqQC7aBmJ zRT7{{zzRl)b|<7U=?5DPVbVOJ)wQc*VPoIa8QX9|6oE^h2KRwjr_yZ!0*?j$t9(+u!C8_ApAu z!gVp)?zi#eyk)kCJ>BMxv>#WpA_|I`a4Ag6A?Xuu&rqI+BoV}8HjXUYXe^bKDM?9< zS)reV5P2OaKp_*bW}a^0%tk3(mX{pRj+b)-BFsY4GPecTTf>5yjxIjtT6vTR0Q6#a zGAHB0c9UKvf!$sD33Z+2y*neR{+iQMJQHd%J{yt{Z5I(cS#NRz{I8L6DKFHmc}mCU z69itZu_wZsZE$}{5Z7ycS`;UFB9+1hBJ}fveno;w!%hf$z=qTpcB#^CbY2iflGWrR zUX;#~S?5`G5#b@(bRuU?xGmUt5rha0TxGnqpzXGoY2enV3q3l?2#X4Pd#|me#f1bt ziO+0>19Vf}UeFzK#;^Ibf_)hSE<1JcjMIt9H&lZrB7VrXXJB5rpq{@**~D@>P9l-8 zKh3W0u7}vIbu~8I5c)s|T02!y0#0s8-dUYY=WvW{;utvXsF2B?gTE$glV%eG_*w-i zVlgTd_RQtIz3w_y@dvKsS(a~bLkG-vYJHe-B25C=J|bKNew)!^A2TXNk*r{|$_#IP zcLwow%P$B{Y|vUt26iH^r zAERNgQ6Jm1#dtF8Xd0(Mr{LRE?j^{ZT5B3x)c%!Lm*9)e_w5vXY?GgwHsNZ$-q&}Bm~%$-EB%~DP~WD$7k za%h{T^WeOZbeXJ>beye_=tmFfQlDdedebx=?2txTJ@L zJ)xt?)?~8{w{MYaR4?+AN9jYUOT&%+`F;-2Q%|!w3Wk_* zb(+jle7(-fvXg;*ZAlK4PzqtXh)z;}BG?2Dlnve`0Fx)Yt{fNN4!~02w|HU#n7`-v zkfH!Q7h8rP4lFV_@iw2q;iw+(&@jM7rx#D02x}|X{OuMnHluaJ=^UER=ceiVrFj26 zDNN5sJ>>@;VBlGK^a38EvU0h_)_CQ*8yPUON5vejlFo@Pe#MU>w!>Cu&0r`_hLOxT&cft+7^R?;)o>pP|NFD~&K^gw;Z!;Cx0bwUYJqGYFy) zr)nf{0Lcc?1Xw%rZiM{)(|LtjIUe@owp@Dbp3~649xjUw3lT7X&LYJ+x^7zo2pq4F zH0O0!q0h_Ez>fXgEzeoU095LxF7|-FgZVKoP8W?Pt3Vcx_$!HlAmOl0?5Nuw2Okk-fJw^< z??_*i1fin_Kno=O=zEwAaK->TWZ1ODjc{v0{fJ;IR%sl20%Nse3-8WKTxV1_#S!!j`Fd9xV~elljYAM z1yNd3BY6Q}-{VP+QE0d6=tXn0rK*Vz=d18aQSxAktpT2j8(4Y~E+??hCS1Y7F*ttk z6u`MjcgdQau`0wj)^Mf~0^QJuF+Lm_omHZ8q8B(T2e31!^eXhOgD*0fzBCrXw`Me7 zQj4qDt6QoY=4?Iaa0HTv9Qefp9A__{LDYQdXG|ionU*Q*qjymVu{&*+2OZ9Vf&S4c zES}h_{D^Gkpkr>iZhn$n&oyg6Zutx+`Ys1+4)Kvdo^_Mq7AfVJt2pQp06%ZlEcNwF z_C})0H?%#oTLZ@%!&<7k22RaY(IW1UTzv%O(waU^On#M`TYZ!Bm9Nd>8uqymN#|bW;IQPdLicvLiEf8yNqJy_6=f0tHAzVuqAlw4X$il^U`ZNEs@X zr=hy17BNp9Ht-5pm-Lt+XG?MD)@}BaK_DTz1Q!M+MJ9B=^W0>u(>9#Z^_1>7+d5YQ z=3jyH)R4Ho)QwHCOHyD#)(8?rq&VHQEU^jDiYX$6lLkHjJaPzthMcwmj*FVtXAml# z&t{P5v;` zXuT*US9Rf>Cr}%CyzeC>d${gVdoa?X~P-kgxeBR zl8xabAoW44qLos5;X%{`&ICJcp0 z0Z7YWmZ`AI6bQP+>XD_n2K)<=uI54vpsei1F|`A+KW|(}q2*hgW~Iw6h19UgHfI30 zCg%r!0q3|;f|rl>h~pOI=Xc=rLfmUn?G6$6`*6;GbJN3cB>3Ys0mucodW>|d(A$}~ zxm96sD`Q6qfan)R(dVDPsBw8#D-!&w@v#2bjoR%W15OnXH$JOB`d^;vk3Ohh`c^!w zuN6PNEQ9|s@v!WF{*F}n5f}jg5pD(50G1O034HUmdmt5l50drC7q=ShPj1)=MOVMQ z6{-#=u$JM$r`m0JdW?hAZ{EIn1j_mO+pl3#fj54GzyW^Z?})n<1?>a=Rr59nM4ua| zP6+?I*WZ6~uO)x`IsVBVo#dgHZ$GC#`9=(`CxUw@%XVt0?}DoIJAE{?2c3F{$M6aU z*aHdkvkAM^KEEq{zSnMjIC;14-_@2s6l4Jl@!L9X&QLYo-uR%Qv>%`W%Hjc+49Hgj zW(2?SCqep{CX-;|Fr5HzXm8E@3VJR)He{aqG{gxL3l4jcbYpU%%=cE@uzR4 z@lPFohA8pD7l6R)Xc+&r;5L?-hKI(li&KFjhAcx z8=!vvZ|SLO@39Yen25jhO$+haFuwVo*GBYre&aob>Q~HcfbD39;_Y2CeeOOL2VH@;hV`tjq_;O^C?=Jy)XN8Z!F_p26I&*th^|68xTgqJ8% zfBaE=J}#m67hXDl*V{+Fdk zue{NhM&t)t_|7=~OuRqCQu=THC8DMIB z5dvD`FR?FwnSc4_i>Awd@yqyg^vmDg!A-+03+8;UOV0RZV7ylsnj z$g=rcnCQ0wSb6E&Yfu+Ie|yI_a-b1Fx$FZ#nst0=?*?fJFp;j~LR)p<6;U2vgBuPd zzz#i`d~lN++(2fVIEw+Q7Jld6lTW_zPVw)8&HQxdll6Fm(0mVuHvbVu_Ds0_D9ynd z|L;GLCa+l*uiPU%V9EZ#Ies>tAF09bJ)%c+3>f!c|88&mS*O426R-5@rSb3^53A^L zCqI6Cx&8j<4_?)`I|c8ovp@e&o;KX$YkcPveA9w``q8ic(l;;auP>aB3&&?&d2R38 z|Is@~`geY(#_v#F-~~Q)(krqO&g?XAKQUmpJqU!GEAJs+*|!)D1kRYd|_Z zw(|DfEpRL|Zfig&y9u(QZi3H#{+507_@Pu>f!U%v^j zXAA$~Gc4FoaIyBGPmuAe@?Ut>>~pt1kwgT}snfxokC{^I)u{@9uO0{;uI ztes~gc=hYo2K2&<`Uxz)$HCnHej3PKLH_1nc=#qc*le=U2%3l4oH}7w| zOP=@lJlg?3`(7I0eT!dx&j&v`lOGYAhaYnW-@I31&qnn8>c_ha%J1=||K#@f%NL&h znB9F1!}tBm)J~siaK82QZ+-dupPv3IMDE_gzu$&`pTNJr{_p-icvg46^O+8L&9wUJ zA4#Cn-}#O2R`r*^h(EbmEjPp9y`l4*9OLaL@z2`RC-3Z|hok`bpHEMH5Oc4ARpY(4 z_P`ACU>n^{w%cz#xkm2;>py{>y`SH&?jPX={Rbb+@ynh0=e{7T%YdF=E3@0VTv?|#fK-*JL(%lSJm%xB~J13dik!GF88_&@!aga0!h?60SJ{_4YT zUg`G71`^Y*-rT$55ZFVzdl&xF9rFNN%bx*_4N7W zO#Xi#H1eHFgSZKtTevMs2v0l$`ag2;+{DFh_3fuBD}43?a{nJRchB}i)*p&9?!ERPo_zo9Pg3lYpWO~F+z$jmEa}7UKM*9{ zMhdo50|8|S65V#)2SiS9wr~yl{8Bi-4U04_hWAF$?pHF8&ldTw=m%ii{5!DFo}YGj zE!gnNJ%8^>M1Al|{;=Ngv&s3R@5?Lm_)<>(u@rl>qHce_-0Q>#ui!gLM}vIY-H)&5 zE*<$FW7)s&;8#CRd#_B;w=cA3UhrS}_C5GCmtTG4##j3AaUJn&n%@i2mpUSP^TiDd z`k6^z+qfuQ9?dhj)p@{1A=9U;Q$}YVp)WE9-vR}v= zIekz&(JgIuOPYZd=fGDPB&u%qAEdl%xX_efyFp5BghQs#rPh0&O8Sk55kA{pk5^y~ z|HBV<)k`h-=KZWZYQyK3UcRh<{6SOSNh|)x|Lp1YJa*--7600IdI`_Q^c^4g*pQPy z`v%Obha-Exc6s0aeL4QqA9I9%>iP0{|3hEBy#7jmUlV+fbN4K}#Qzv8@>#!roc6x= zc%Io6><``^&xZ3OH1y-X7( z_h9`$c`syMot6LNX0YB4rWY_d0f_HY3|zPO!zoXP4qi?M%u@{V98Lyw9-LePf3%6= z+`-);FTn8z&b@ah^KaAnNR$1FTMP0$hjRnK1AOz*A1EI0t=woRG z{9luAu%#cPx$gqDuoS=J-A`jm{^GxJmr=iU``@#Yly9|#?lS5>f34pyU4#GZ&GHw& z{K5~?*7@>N=d;i69_Sapgk$7Uod43PYiIoO=`1V$rSrSL#J~LO;L_fF-J31@<(I$b z{2n}1y?top?N7h>;_=T<;j5qBAAWl~35$PuyXA{Fz+Cy+PaoU3i(k>?uTT@3g$1IG=yv7Eoa}X7E2q zq}&{`XBVCvvM)cmv%23Sy6@t*H^rL)ndmrs7?2PM4v;70BVDyV5Ye*_H$w8d||&IM<`EYdMAvYTgf>)qoCyl0o+8xC;C zjeg>0a{&6*-Lq-D80yG>_P|T~6esxC@ALOe%)Pq+0s{YFqrLQBKKss7{^;I(k0)NP zq`&n+N4|B(C+|+8{kMPh*B>MPj{~`TKd+rpc)a16iT|}jDX-k~Z^CPTZ6!UM&2PNr zmHB;5J-`3eXR59GLtm-SM)XJU<3D7s{>0N3dp3q=7hh@n#XA31r}gw>F3ju?R{%V1 zv9JD+fB2Od|G4S#Y&XBx1b-l2Zv@VvJyL6b1Bdq`$2XLHlp`cqfi(N^(v4@f$Ue^T zZNz>Qyu1x^YjA|yyK#Pl2l1{8`jbpwygSMK!9#Hw+*HKl{h4zPeo?-=`0~~VP4ewj zbODr`QMn|HSr>D-?1^@9=;I?=RXAl zAT8?8{|0!NH~RAX?Z3Ufog=$H^Nu>Y@optM^_^Mz@vgtY1Is2n^R9v7KF9F6`5dDj z$f(mG0?Po#dSj+;;K)Dx3Fb=p5FZ75^>4np@paz}(fDT0j1Ql0*C2xW`KRO0#{4s6 z%YfAA5X=o=Z3|NSRReUJek6O?_E28KnCQ0$%%pS z*)PyvDKYTRmfu1QY~Of*dGdi6n6-C+GJAuh`6D6*#*$qsY%GJ!*MuL~ryUmNEFI9Z z^A#25d}9g=6K&3pB*}ofB+0TAmpek`&7!lG#J*_Tg+jo|p0;lHach!o(al9GYq*`o zvB#sYX^|>On~ek*w-b+`5JJcfr&W4h)xcTSAXi#6Mv9*L9v(_-JPaljb-i#gzdKju zM3fakV%qJ_1%PxPiSlU5n@jDbmQ?XsI#_5`uH`Lfn^{Vx+`I&&Ayh6cnORdF5bA2E zT9$i7j!(OqIkm|?2NK1xng;-Km|NK-TxREND(-g4Vqq+gWQQVK6_8mM2M_J$U21Vx zla!}BARR8`T#dh_wpU2Hvt3|jOj5Zr<}185RA+xKZ$ROg$_B^zYCckjmN7a5 ze!p`7rVwBg8ACTP%Piz{z2$phIrym#EKuoW$I>C@EqvLrW2_94gU*;0e*qS^%V;fF zy{1Km&wD@+x2M1~m17G{FZY1w37Dpnz-owkq^^yR&e{=3n0g|g@yx~WLrS-88wsf{ z!0_9G< z3lOZjZHsW>5RxoEIgs$1GmDuge9(sI`alCk!XCS1+$jE1(1*oUZA~X#vnN*P*oO7XO@;v_CK|dfEps=jBOY{PCd6BA zlHgLf&ZvGURF9z}>Er>fDGv$Zq#L+(H}i3JN-n#)3F+-_+YI~5(9JJix8mUiZ=8*r z5;JD%g~*Hw^%+GCL#Wr6>11lG&P0M?R!{>8B~5!_iMwi!gf#}Bvio|&0-Gzq59&)n z7`TRTE6W_G)JdB%0ie2WFB9DL`)o&bz$9><4{M+p!!kPHmoVq4nqNrmVo@k4arnh(Zg=yI!B7BC&MX9vwp^p^0)}1B~X~1 z&l`s6R-~q1$Th{W`E+?Scb6&I)V)1nfba+n!#+n{e|^aS+4(})VMvE{T@$wj5nj-p z7h-OLTuyC&)D<8%8NBsTG0{}^ww<@Sj1fT83bQqE4J#==r8R0c!|YJQ2}fH|Bw3ud zN!9J*SS;QBQYuQ;?*RgOR(U){0>2ATcYA>D2RtniJL6SAgr~9DVH*Pu;02urtg-<@!b z$^F^V>GgrV9HYYkIBNN78m=w3uBVx}H2bh4j^`^d&MYSv$rQ30ZL{7^l+G5j`dEM* z5>BO~9MxqnHVpv_sNnX4I7x~&o$MW~SGzEY0vDKUaod-N!B z=U}}Qb9N7MP%Ju;>q8{2B>T|Y zDoqN@2&ktml4}ZXsDZ5c=pabQUTIz~#96|NA*KaVz8bt8Fk~hL>S~0NW&3dIUtRf# z)KmFn&kWIAsHVFFLLT7K?o`?amIiBvhr@-u0syh>?056z;1^B>vV9>%v(_~Ga}C5` z6Yijj{&c^i(?D7wsLVw~XU2J2UpReo+NokYTDztU0q}Z%T+o~IIRm&|k?$_vl0S__ zo=pKYyv6->vMIuvpnGL3_{-AR?H1?J*GpJGb-yi9(gSW8cNl?=%Oa<1@fxm4Iyq7v zEDzfA{lQG_T~!8!K>|)<6l(=$&DZUXcnY)EGhI)j=xUT`Va#SToS5`8c?4)&AzIBQ z&T*Z4RiNALp~Xpm^zs$84s)8`0Y%HU*z;#<9c<6R{6bi}VzDQq904TznVZiJ{z!U0 z=!Dz5VkY5aGM&}*0qr9oTIe^CcJ=IN2>`+vf`a7IVS_hr-M4^xu{KC7Ss0#8Gb=?w zmQ6$7c(Vp|qjk7zoS7gohHzPi5uaYEgab0pq#`Dvo}a8i|_l3C6?<8YnkU~AVV!rwCDiEM+ve}F!H%BRhpQ1tF9(`=DgV{i<*ls zn|iJHZ8NMY7uXA3?YyXmNn#YU<h??B2+Zk=u4h z%LGdd$OjrwlrgY5P6In9T%bS#`J0HziBNK6hOVZ_avea}g^4L|43_)_!SQ8F#F!^rbfmd>?32iOXiU;JI0w+g86P+%~&<9Vg}}mL?0+VPT{>No=OL3v2+foNT0>Mk0NlR|I0W(#1>;V4wgYXfVMs=s0;^&-@P2J&$HAg}VARm?+DfKzHVp zVHPV`dSDgI7d)Ds@@`?O18k$A#JfKBXgkd$;MWq!H_5HE-oEx+ovuN zHKqyz+)I;wSBpKM86a&C!2GT*l?ITwvD|~Mng~{&lP(nk__=Wg5Oo3ty4{m#)TTUI z=k4?)D~1`LEj1Aaa;o-^wn;U^41+J`=^)LB?J+z73z41r5!-WfQowd*bspjZ=(-8a za0!;U#KWA=D-t-OhUH`)v>eU#Ysy(Kn@_^}~81sL>W5 z)od@S<8~=hiqR3=KoS1Z9Z>zw`H)T*DrlQVs52 zMgZ7QS3G~kb`-9`{$GO~p=H^k0wQiXm{^iLF@g7%KkL}H3Z!i*pI<3vudL4V&3wB;88Gye#ZmNgl^~|R zjLc-m+Q}O1&}1nla-IW$Ro4y#7|sGw4GWpBaI^Fj3yby}V03Zq>0~}#8QX2MJb4P& z1GAq4!1jE<1R7W&T^&GgD2L&wlCbG1@9O#mn&)7f1SN$Pof2P(uA~gJFg9+3D$}AW zaLn2SHHvBnf7%q4JU5SO)_~G?x(mLhz=R#BLI+(ruYe#gpiY%Q_+T^6{JQf&emb9gRI3g%cOC4$2hZ zpm7_>fVV-FGD`(0*`skKB-jDDwk2blg=Hcva5`vU0F~zMyu$r02r$A08GY_ZMq4RC zFbC4LIT$5ET+awx1d3G#cemx4uwo)Xr|nMZ7;Fu;JpEUbiAj0?CGL49jiU4bdfAg(*E4|XZHSugoGip)excmW(g zjy12cHLpDB$_Zg>+ua^aKWc5Kg|BiP#>i$=9`{yexW&;-fP9mJ8Nj=FkgZMkknffvpA-2YhMhvQU9L#}$lC^rkA%|h(6Tg4LGh5VP)s#u2OiL$JGD9NItB%H zYcR4`t3}E7n?)D_aRtaH%k<`pbwE(QIRe^x>9I!uiO;2O%>ko@dRPH-73Y9F7F9KX znIQ!Y1UQFWb#W@oC$q}M3)m<6?m{8e$v>XyTTH|01p!+PZLOJn(S`ddkPB_Q{eHLW z#S+Ze>3)%sVp(F%>|nVrZ%XS>YvXBlN@c;BoUlnGrq=Q@g8rD2puX(uX<3{d@{~^b z4wEXqhI!WTMRoD~*eItORYBvRK@e}Dw&VNjObJfa?RpS+bU(prT4Ee2+1PyteJ2-d zG4Q5FdBtGv0(o=aH@cLaPT>`NG0>MvaM8JV3Cu&OrO4Uk6kLTh^kF3jhR(-CTHL}K z@}7=$jxB{nAy-GkBpPE_J2pkdd^nu#Aty+}7>p=46aGUlGL%xN7RmI>J``cq#_m1CG@oqLO?S;MF= zBYb&^T%rR&acqKL2G$39ROVuNxW`^hs@wyGPh66R{MPwow?XM#D(#hK&qY%KI2u=DAYk{noRTm!tE$UQsX?f*4}o*D zWY{u%TKn!38j44xrx%rn>j30KFc4sy6K@H~!LN66%rk@B| zRD=Ebs1-m}LjyY*^mEuJmpRag-Q=(9i4NZIFFM~(XT<*G0QW`eM7@Zyt~hWQF|50OpjEw@Tw%W@Ks>T?G@ywBK3%=oua$N^ zN3l(R*BeFZH41&YfZ;MAx;D&aXtU(^T;nkG2I;fP8z=6bm)L82->0Wz1FE>)FD3>_ zaZ8fCFsC8fPCQ2Ux64z)SwMc%En9qWH+7oI`&AV|Ah%pt064$C9PAaB;K`O80JJ?# zy945?B_h^)IWKc%<~vBe;y3?$CA( z&W54Qcb8fqj9W~1Ud^Ux{P3Iml20lpjIQTxJ>;1FfooxJDHpk^1NcA1g(IT zx$Y!lj>c2eI_HO&mQBzM?P_5t5SMb$m@XaZG)GqD%9MM(Hh~;8Jyoi{zcO2#0alD9 z-B$%r;&pBxl0L*AxxhE}NWVXf3Nd0a)$8lyIX949HteU$UbXay1oZ}F)U${i!6so? zz8&*3Wa!b_~O!#zpk_G6%++t83K2n&)|aE}bK!)hX38S2#k~9K(0Bpzg0|J1OL=J_qLj z#Pj-h2jh8?1)~4i~cuP(wx{$UEBOZDP5ObRT2;Yf2u5fkt>v%{*lS zyxB?AjUhQ>b5-wgHrybkWA!svNC!`_4qk?E7sUgX+&mUyXU}HY;ITk)vKaDx7d9Td zjJB2m#*{s68U_NA4L>D?;8L75Hf>$EK`av}P&Biofr4+QP?fg!i6ulyo9i}!yaF4x za#sb2E_yBqKO8 zg!alalTVEu;Vp6ki zv72WZ&Vx@dC7t9#Ae~V`>@2dDv?qZpmg&~9szpFf86}&WjEqnCX`A&8KKT4;vhsS0n$8h|Y)RqDAP;Kc*?;yLYOk#JREy(F8&y?T>A^8jmhoJEM$ z>t&8#F*U~I&D`-YAC}kwRUDBxL{>%$bJ&ttgvqC+>7E*KsZG{fE_>85#;z#1euI|m zY=WZnWr1DT$kW&9cH&vXo^743D!V1~+@zHHYYMa3G+)nTICb+inDFZ?e{gal~lgc=qSP z8ggXmte)^ybG@t&r)~=psxD0_X&n?hAKk;i$IaZ`&981hqcegfJszk|Lz<~WRkzhm z!XOumnbJNm(l3wAl4&o-s&=*}#5r-!ZTe-)_#wAUCei`1^d={+DsW_nSK9~LUPMLx z>k)9E03&@1%>XxNAcS=#Jcp=(%SC9?!G{1SCL{( z7`E~@&GqI4?}Evr^=i7k;0&Uv-JV957Azaf>Um8LI!uxM%fgu{DVNC0GcbP)jBhng zPMQjWrK_A_sPz#9LIjazwv!rcag?UOY%d#$DUBNeY|tBXvY*G3@_ejS&{?L`Y>Eb$ zd2u>dW%p7pry0>5h}jOA;^CS)2rE!k<4~Q6D0IDia3>B4(PN)=arUHgCy?c{Vm`-4 zM4sC5#)#|;j|a$v9d0Q-H47%!xUq>duOP%s%A|gstb4kPJ?&a&ipgt{z_aUt1IM#8 zl@aLSHm%>);IAG=oB=*c+3*E?usq+aDSz9gsYR^<2>xw#VscPzgyHw-xQi8d36WH; z!+cR2^dA;C>i0{s#K)7}>H9aKCYdU+m61qV*RaW0opHksCZ zUi#sw*VR&XkaO;Y;Zak<89qTH1D(&orrv-V&h2#-XzVXeWkp~Ui0XF%v!5`TL1*}UAMDKehFpQM*O84v zP1a;~;z>c|2#E77s3;|4kAygus736RwMD2A*0?PxRgxuFUM!Fj zhH5`iPh3u1*gfWG1ix4cuxN!#q|PEu;Pk== z|3aCo8ZRyB!y)a7i5C`=Frk+j4%yuQpS?Gabu3Ho!@6gmH{;Rf=^0H+7N@49dR2V& z$ji&MH6@ArzVGbrdAcMo$;;)EybHHMC$fM9viy`-|JQdivw?pvvy-_fe#7$M4uVlFvEO=I_P<&QjEMW%7 z#xf9Dw#7xH*^70RbTXUG&IG8DT%}E>nQsS!qj(h=8aG8OkY0a$_YNuy8!@86X2 zyMO`os1QCd(FtXLLes0M~S>s9mJ-YU1#Vj7uZl-W`^lNC67F$PLD4eYPBo9L}EP( zmZr#K3j>Q_AU9d`qhfHuJLp?9%E5fTz4rO!egL`q z@)cj(8C^DJRZqn^C@lC@DluvLaPQ}aVWtGx-g4p23SxU!K?1F5O0%lgGK$F5*j;afSK;SF$&r?Lr`1H0+K zS?w>qazB<%DCKkpx!MJbt`3}<|==;}ai3w3 zTWmY4?ruF7TGXXUYNdu7a;(h6lyqY0&8is&oQr#q7ly?aKW)!>m#@`|q$cCB&y_4z z!|H4r*o0lm#uXe4e8|q&;fIi<_*_VPpC4tbv$dksioVdQCWW2BYTngMhMTdee*r$X zw`{CN-PULz`=Ttc1uN3i~q5`87R9l3+f(4<*{#w^iBmhPQc zbT;EmM+pyo#b#^XTw&l~P%_QLXfjUw#C&T~-3_B^c;k(DD{(#|c*EURZW)Zk))|J# z{u<#zyxb1&Hmj&BMpV*WHxo<9Yr*YoLUukXRuVi&9p9!J}BGrm;9o08zJGYxrh8k)_uc_162m-xXj&-wUr70Jrn2J>x{ z5+jAH4`ROhdeHaHO4FHVR2K&;xu1i42no%LE~n4a7|z7%rj(DR*=?n;%?ccHMi)1p z(Q;ayjG+9jI@z$>1M`_)61hx;=qktzwjL$d-$jG~-;-VZa?pvC`y*amjXYw_zi>^K z{$$9L7^}hhpt6oOSYkUONR=`(+Xa%TQ8v5T1{I;&@@@L6uiEc6#-)|uoVwYJyR=xR zOCa?E?5d_XD-tJo*=b62))@r`kib8pQm^sRd9SAU`mp%fo&otaRqI~%SF6KTgxV)?;x1wPSNPdhW~c1RtDKxJv$xpo%VqDh06Irb~)opzGl7LD+DI6+mwdtCexa% z{pm`t80m6Mq#|r~*&XXCg?9Z$P|b;+?YIqLUI3D$(#*#@%S^63Zr#Y*r` zhlyIM;>vo27}05zmyKS+lXkM{da5+!hLKDvFBP5LU?EpU7I`j{EI0!J*EADyhN{ru zdKW`y3VO;>%Rg!*^X>MgIg`t=81Dnkxb)`aazk2n)03Xx1v8prc3mI$B3mZkYQdII z*my^Bbv*J%mVPJfMV5n-+#AIG{$bKzY-Co6FtK%-SF+IhCj&8GYwch81G7%6C30M1 z;~LjyvtDOa%5($8Mo}%b{b~v9RDs1!$tSbUagxh6G9~C1q(#|PDpuL#pc?Tp{#9-> z5MYlNYe}pWj|Nw)>=m({)j>U&Yfafu#kG`M%~HY30@b>prZ0{H!$A+$UQ7R+9~U-E zc3SOhr;~7oFYoH@ewb+>vFmanhZx4S2=2V8-V|n}U0`GgSKF>RCRd^WoM!0NiW(k*u+JZn`G(v9(%Vl zXnzGiq^`wODBVPg^~Nff7sd?V@@2AvR&d-Fdum&5rPW-wPo7ur$n` zvNm;%T)t2jNytm7LS|cTB_LPaP;#Z^nlv~QM zQ#f%An%qR1hC=}X)ImVy{fSIt=L|)2p4BQ}=|LPR=IgtGoJvQ!g>+I}R|chaXApCj zYO8r_Rd6nSizOnCZ5w`hm8~s&aep0Y&u6Q-v(7If>Y!KQ${}~b+m@M3JU6e{4%Sk93Kxa|pgiRArQt3t z%}a?D3`TOZPDu5Ywytuo14D^Mn;C>gP6&V8MX53Ea=vc2J5S_Ae;~yN<ygA9U;bt(_Dl?!X!h?1+-x&q# zY^vZ6Wp?r0HpdmbDp)VIJqz>(ohUn=2LxA-g;B)LJ5H$VpqJdSLATgrdukyGL6+Aa z3-b99EI3lTVp=I}cJeBhWP3YtROnVBWoc3DY$60vs7S_+Y}LGXyBY18g>`cmYbBGB z)Mi%dRVHO5@=iw}&IvNWaVbcF=gn~1%PGt#60BrJmr$E8fqb`2uV2rtoIuXSic`ro zTb>Xbnz-8DSghdR2nz41`AaRo5Uh+qpq4f+r!x|P38Dv#RLdno@$66ML)#7v%+?ua zPH~FJF+PNWW2IUyJHulyvsvvL3eN<*@yvJ^0hPbxbXoo@KJBUM`w;4%V$zBApm{y&?>N3v5OT zOndp}pqI$jn+bQeT^L8?RLLcG#00DXy8%f7G`<6ZB< z-&+MX)snB~f;A#U8Iy?W$rU-S1{CvTp0H%t?qJx@heN=AgH$*+i@3vi=V+SZBa!fCrSc+I z;MY!HF;WP6c1>g?A9~&Sn6nfsr<1tA8e67coea9JU@HX!_Dx|VZ=>)7NP9|A?F7?|70Tlb>rNoago1^vx93ZShF~%ImNc0LOU31&zl}=%9h;4e zSJh6X(Dm;ktzKOntTV}KEdb$TCMc&ptVay@hS>=}N)$M6GqT9ZYiX5cOHr4y?x{Y0 z1R4_pVRjYIbq6gGMq=Jj!s+is3wtGLJTP2^! z#4~P~nw2+vJni4A&52vhuI4!F>)tTfZg7)E*VD|my#8geJ+H0`Tr!37>wLPw!?bLY zcX5TrYO!)o<(%SiPP1aSwdEKoysf44!)e)*@^+H)XdL$TSD-2LrND3)s*lIXTFI@r z!m0dbmD=+8YOWpGsU(#d^9~!fxSZLHt z)uKIk8Wfz9kQ4qZzGZdobdOxA(6$M8k3z_m%C(ZBq%le7Q6q z)EBD-Q!>Ys%krI$CtB<%lR&G|D=$i87wqhi?tGk!cxLMclSp`6F^?+eAel)>(?z5& zx?tJ9gG@KxnlW6}>1uHNY%||+v0kIU_4^sNTJ?rT-B@=%oBQ*Tyf^A^W+6Ft&sQ*o zj)ej#t}Jt@^j46PJs4)TT-}Y3>UY@$yN&dq!N@o(vMe&=T``fU&z7@XpBvBX(#pAt zFL2Yd3=^jDWWZO#sg&hjG1T{p{n^qLU$t_H z88Z+0GC8NrKqTLePMa=h5JJ=COmqgOId7o9OL&;wqF9`+A@GDHdU4%qg`=(5Jdw)< z8&3aHY`Vo6?-^z0wGiK07bV35tBm@nx=~WmWXZVwYw|F-4EbP@m{_bkQZWFm!AE|7yn3ITIq%xv5qcS&rghox*& z>V^wyC9JN(VO32>;{k<@G-SRC6rZ0D7TbKcxPeV)w9yDxp*QmQGs9tqbT(-9XhFfJjg|gtT_Clk;HZV&$2kz{)zuvAn z_`h}7aIP|J7W=tQuT-h`F(7eS>ek)kezVtJIJ>KbC^`F!St}KejtV<}%q!*m=y2Kc zrShq!SR5^ON;J9Mg`&N-Kj)U_g<`r?DYI%;8V6m?;*xVYtAoVQ4arY1x=q(R3rHgt zozkS_a!#0v%*I`uG|CiNshD!Zfh=8^js}CeT66E9q!AaH%vvZE*;+(V3NCMihk**r z=F0wgX|Z#~+dVK%b1oHWt@(nYDw}3F;D(=4aFLkzD*c(O*lC8;s29wO-1Id)f?x3o z#Yjt8^j#I8j$&z%?N>b62_t4QNx|(6aQ)mgnX8G;;?mV`G#iPmKe}C(Ql3pp=0^QN zUGVT+Hqt6{xkg@IK=zU@F_BHXk(fxs$TYQUtMRVfTW4!*C&ai_C(H4>TzpX9r2?a9 z*XP{L7IV-)oAzL_NdzRWQmk#8&Th4z?9{4bWfu2KNuPhVt%;#+Y`nLk zkp`Val~uXKU=!>2lG41=56?a8;$+yZ#r(7RV$8A|dC`rh;GWkQFoF!5*GVIsX~4R2 ztyGKMFgpP=?(YbpNg`7&&sU+5XSz{}?Nqs$;3C7OG82pLR%Pjl&D%4v9jLg}T4<5$ zEuarcaS3L(EDW4JH3r)}SAP=V#Ih^TGq}f_YArku2RTJrLa85vzf{Ci7J%gHU1u_- zM(SK}T*+)_P!+_Ql89JBrOql2(;}(e&ljptI_4t@&{uOV5$tZJsnD`flgGi1kM(w= zhH)qTr<2jbw6MqaWR_J;|-S>Q<%XY!hPhW!2?U z`qQAY>~-bwXftSWlTp~sM3dd|ELK}=SaAgKY-XUpE>!)qp&M3dQg*c(s@2xqzg<-Q za#UW&ix6w96i)8s!>L(oJqb*M`AXQ`2*(@cx<{Ber}0gqUCqv_xp1pkA39s{#IjeN zc-l;JoC(L3Vtv~wbyEFWL#;7^vNCFTxO%_6jrAsS#@}dUdu%tgY!uXW!M!VtQpJ!G z86>iSLBm-b4xIs@?UF~Bb&^VOxGSYvu(MWI{F2G1{mIy_x@gsuP%m3=Op7aNF<4b+ zyU?m2?FMtc6Yor0%2IW0+_3+3cctJm!fkPtrJ-`ORS{ORRAWv$1Wz%Gb)^ z?GuslNRl_c^(4{Y%gv&9q?EJuj8LgSmYH!$wOJxo%lP6Ee?2(g#l^Ts@g$^x5QXV$ zqW}wnT2U&NWAHywlB?FVR~?BySG&o!!5k=Jpg)0)Jftb`_3`wT9g|(SBsNeeId_~K z1^WXKT;^tA^ZbLI(uzeBP>-cNllnGQ3B^;jYPudGrnr4r2js)GbU5HmY+#OD5Xaog zorrh6kJ87>EgJRiq<`Bvgv6>n$c_*8FeSEaZ-jQuNY~l zQoNx<8XpEKnd}nmfSiIJcfnu^jc#G*n&j9*q)}u_Y&gh(FKp%} zsjw$B2=3VFGP~lu)h*juEp{=sQ|`krubPKBL2%w}G5H2N%!~)El9b4Lf&3QnFx!Wt zbA^Lh&6=seqoCgmL$91(Ckv36{%;q5;>yM`)QRdx+P$&YKP8jMCf1(lyRTc!T%*L@!Y91S; zmfKw%SbiAKrmCK-7_SNSOd}H;sxvkvl%Yc(?b^#oOHD zype;|V&4_t=3T3J)W`F_DYsZP!^vh5R#l6hG+eZrP)mCG%qq9wyuQ{X%e3dtRkyi< zncCDbu^tLjiQgt00Mu2@=S#QDT36*xRaPE|RuqF5KiBP?_2+K@h zxMeF*dB^NpwRz1w8!KxSjBKvk#JagwFiTO{&}gZqlANpNiDxFeEt_qO;O{b28oQl4 z;`ip3?d*2w5}5UHmzb@%LC6cYSz`0M7REC1iK=Cr=~_7L+*Fr|LB4@e!k)g;2&OBm zcqhH{C1DlX$oE^s&$U%ybM7co*`yZ@>gy`ZVP;Qku_xX|UW+eKPgL?w*w06L(IPK) zqzJ#p!WM;kGw9IqG=^M%&K7$EcTP;z zpaLg8y#Y^l0S9IHnk-Vr*eeZYo$$ezZ4y-7c*hFE7v$glJ86ODlm zbcK8j{AZGBDU}PcfnYGMy7Q4r!dI!+JVI~i&$;-V5ZjIU8QT;l!-2RCspIsF8!+Qq z*b|J^hRG@fKVgND0=003?-xAIlsJ^!s=ElF7(}vFUw0MUwq)R77liViUL+UJHp4ae z3yD40D5>f?Iv=`Sjlj$;uto1qEDcNE1rL4iHq6p?R+CbAH4pm(yWU_YuA}{~Fz0*W zmP_#ml4D72^j1>Rlh2m52E)zR>CR=7 zxqP~q7^iYVG0b?GX|f;bRSFVdXETpF{V>7l;)uBiZTEUw4_Do>_)03v@RMeHY)X}D zrADC6yZox;F9X<9u4r=*Tg%Z_$mMf;Q{{>ivgqktOv7iL$w`%9anMT*J4!qg4oTzo zXfWdXi+V3LP!nOLlUA5zAzYbs)Nra8jaGBLX)9if!`HEqj0{GLNhKSec9yIr6Alm6Q7@v@We=}%GG29BKs_J^3^N_K#K7hZI5kmb*6?7uUbwpX;?_Hoa3$KT1p(j*Yy=X9pl-@!o8rliV#oQO! z4!0Y1xAAw9l@(k=e7v;oEZ34BN2uG9+x~U1BgOijf)s8z!)K%(?M#tid1DqDW|+KY3Ov%==ov8u(Dk)E1g(l&G~EXRItAr zL`v;2tcM0WxeZJBBCA5;{IBEH<3IcU>G-eV;qqg$f-7kD{`hykL;KZ7nDh6?ze|377BTwV z-q7Bkz+Suj_-~+*@h_3TzE1u!8~wNT4#G)c;7_LghdiiN|kL>Cf zG)u<&*$*Kz>-aCBmC@sjEtb^JnX#k}Vp094D9RmVb{+pTssHBK2tYM@Vg$0<5B@m* zdxw=TKiPvr=7q9d@b6!Kq&M&LBvb5I)Y>zmL0auu2}oMMj`+Yiq-{g6UJ^*RPqNnw zBC>}#M72i(;W`#0fNW2u0d*yQMq<;?RV1#Tu8aSTAYjl3S02-pHZnyGA-g8IgngW*QV^Dp5tmx>qukdVg9sm=I9`m3IYg|aKs|7B5opL0;3QOIecO-ju@6C-@wE0 zLpsOah@?w#JnYa6E2yk}MpLvIlFAN!J}hXTk&ThVf0!PR!OHv!$x^0Rhy_9%!*mRT z?URHnyhK8+BWC}4E=~Ju0YQ+p=^`BhmL02)CD-Syg-14*1s_I8=|mFG^#l^8-*HMp zCnJC|20a6@LSCU?@v#+UfBcQxQd`W}Z)C4HC%<6^eACh<9}5nAq|A@JEDYLa{PJzx zc^(E%z-%Ah<}lbsd%N;SMzk01_wwWV@iKx|o4@x1`TfI#?b*kGhqzA(c;WhQKq z2cRCvjjhTmNkxR@XkC)=Ov^5WxWhA=34l-yl$PK(BpkJwo1?om;O0J?GTZq$If&By?Up#pyMzVJ}bo`B2(dOe*hZ?rBYTu%yIb?cfiWu)pCbIaw zzi489YXY_{{Gy&IPkm^ReEOc|3CUL%1%>VmN$hAOm@~&{zQNL794ze~TE2W;HTD-z zW_>!x(Lu#q-DiLJH_cfY9n<^l^H)y7?DX>LGDoIy^BJ)X#aiHf2q~PX0E+zSSH8+1 z1AtR|?KF>GabM?OB zv2wE9ceCaa(S7-|FAsYk`aN@(tfjHR-W%GL1bzslHbDy-4lzLGs?X*)$A6K)&YR<( zz6%J^WBr|<{HT$l!y4KCroDMGyI7F3%zz6HcmjSeys6mJ&|}p)(ASnRhe?<7-+1}u zeBeGZXxnmr7#RYLwWH7G2*&?`wot!t=Y{e#kMy~}P)RJ5NZ~ki?(_Rt|7q~?4KR$- zWL1NHzcE?*Pz7 ztVuIe?tgI%p#LD6Hi78!gJ>N~^;-u^b@lx>ie8rP8w!X(3m*B8GnJn=VcXa|dZRv< zjmY?F%UHy|2qakx`$E-^0NKrVwQZ6X+P59xzi(bh5YXweH3yYzYzbjuWs;-ET>l{P zOI08uHAs}Y5T;H})CDqU4I%OqwQ~b86qWsvY_&sQZGviMlx(c zN`_n*a|MNIddWQk+Qv8I!nCAv1ob>OXp*c0NTBb~No+#M#qVXBwxfSf4PsCK=-@OQ zlO`M~!`ykvZKHpI%11`JNAQlHka_5Gh(k{Jnb;`#chRguKQe;Hz-zl?gZQBcU6fa z)4BfO*u-#l1YzN1``aU}-;8Q)@M@vjeZtXo5{YXX1<-??{6)$E_6BL~3xu8|k53Ig zn*6Y+?mO2rr$q_DGWP5xnd&5>A1)0|L* zk2xPLsw+p%PYA(vl^EqH=GSi^8{WT`f#(iDc@Ljmei&0R$SnH2zoz4Z%2yLHM1iIY zrKSJFTt%aEiqd|-VB)V|GTXVhZ4E%O(UU>G98442U`~@f?@#l+`<&+Y9Gj-G-B!n! z*OxmorDI3`aM~#I)=NN?xhyZ$%8{|(j96??W1YQy#PVzpv1rmc3%lsQbrhrh($LU8 znt91q%q7y;|CdKPKmLV}JukDcA)SbhboS=^%n>;E1ViqJ4#3560yB1UY!oexnKSG~9+P`tAuK2pYSigMG!L6{#*TUxV7z3o(F0D&Pl4qY4VKYnZ_>Vyf%ibu$BG2&ViHH0s=pS0rJ_JBYl&rn84Se~;9G^#VzF zY2-X6k_VU$oEb<`o(CqXr-%MKpNAXU)Q){UGWP3N*&?AwAKO3e#z;(batdQ0Q~vcG z&jGK2N{)On8@@d95kvmA&)p!Jk z|DDHRE4T|L{FJ-!gun7tkBfZ|kL)`GiR*o}^A*o~U&UJnZ^>(~_WHE&wDXvLde34w z?yo^}y(m%e7`yk9oXB4vnYW&`95(#@<2q3QBP>|8hIj^dU}8n>$cT;?Nx;F0=u#z!x{v}F*nLjxT1Htw@IXw?Z7h_^+DlX7o%Ehw~20WWy+Tss^|>`ZQ6*55sQ7L=hjx`Wr}rKJSA~?LH@b zez$$L&H1i1k4*a*hCN&wl=<$fVar^Z>k1v|`DR38gOH;q>)qNM1`iS;fH;MMK8!W- z!=f-8Yu+*9$afjJjtM~opG5MK5J@DRHmzmRCbZZ>0c|3`Wo-Wk+>5n*A8<$5g4JEX zmTGB#**c`w7L5PFN35d)GzuZ>>G%(xduQRmRuR`;jq18U7jll(e2U$m!W4q(7@$T- z8W(ektUQp`T2MxGXO3XVQ%X@0V_T{a;HENw0A`{ag%@Q43Wj;<<8w!IL+zDp2}JWi&2Q1soVI3mjc-swC|nP)NTw<%h0{M*JHD_sJaJ_az{HDA2<@!zpduKWY)YM zf!6jBC>#g|Tn{&$h6f2PND!$(W9#<))@|l6Pz1gII1^v#*QU`NFqQsI)0e{HzegaQMVafTBqgA?nDLVm!1($XeiHjF>23_k<1j*s z3T|rR41wWNjLh|@a|&%1Sd$=E9z->9hZw@rgWUJ%kl^zp_|xbBWVDSk{yrNPHA>ir zV8f<276n%Q`T<(p$EqjRRvEa}{>nVh6vO`7K~uNFu)k_ySYs#@!#)Na={zQ(h$yM^ z3OGP=827w>ur%N>H=^}w-R^X#o<9IL<5L^R{)Z-3ydKzfd}^zCOMH6yuh@=zbgoY4 zv+f}>Jtn2l5i5`MbD-0D4vB|gmU=FAc?DroL5C8$aKQ;cl>>r)i2Vq`juu|sA}0cb z00iY3n9w5T?6~FFxUN7$M~eat{q6zo*hfQe92dnHDCw%t=Bc3Q>3==w;`$xa^^Skb zLQlpJDevFZaMS&-uQ}$+53GwL%Y8G@+ZN(~z8lcnYTgp)TOZmZA_!Zwoe|XlTm^Ij z;&~Aqg?Mtuk`nkAG{w9pP_F`$ODl2eQ~(Foo2v9IEF?kf35%(A4dZGi$Awr9E)<8`d{wGvw~qTh{)u zH!NFc?M({@Jhv9Mb9QYpZ2G6adtU&&f5!gwF}bF3!j6A5VU?b*&mJ^?tMm1@t@C9J zkDjmpMkliVIoeHgwJ!fToZJ!cx*4a~Hvf}4@Z6FyvDLJ2%C#Mif9E-t#*!0R0-($P zstK7J?%BRW6WH|e0AFH5p-3)Fq8lV|d8UiG39>;Q3C_V3BT$$nsN;l_1mdSKkl^!7 z5C9}UeJ~ePyrZ9Z(xAq67|2C?qs&j{y`CzaJHnc4u}w%re2)dl;i(riMLv+%cY(zy zrXh?DMy<(oEF_Y>$BEgH&b>AZD0=GviyRaC9x3zerIy+@pODTb z+*5-`=5*~*qw?6lVrqOc>U3qA@|A7orVN57f# z9RDbRnK#E@!9O?PkI}b5s_}QNp}kG3*pB}{*1h&|pd?-93GSQ3O`x1Qd?*gbcx5U76S1>L<$IEK*kCpX-*#~%AwT=p`ncsAv_TNS*1A4SY-=Gcdb=w5)t;svj%7z6t-!v#N57gy1ZwyCP?q*!`>-z_K^!n(- zAR2CC5zjgVI#pvNAUh1`LRL#3`9**yzzG$C~;+(Gvyj9U^}a zb&dA*@K{5o0}&!Zv;vDzLSlvP<`6K#07@k{H)Snwa@U9+g0qbS3OlHQIwdSPf@Gp~ zoMK{sM4Ug2%6ahYP^CL^2nr*T(Ss=(&HIo@rignMnj6e}q!TE0h!RP+Xi`N=m~c}; zp`^!?Cvp!d{gkI9xQqyUD2iuc-_-i_tCb=)A=&TgZy3h|l8B+rEmu82P)L6yV{ zt}8>CuJGDtW0Z6Gw0iEpDOT@7=<78k9`qhA6@7B!xgc7-o{^O%WwkcQQfp1%>G~%0&dE>duD$V} z!(YU#V1s{g^ZufFwHtiT{^dUYDU^FIG^0!^Bedh!O>4|wurA@R;p6@#JpRk(tMSj_ z@!j{Wh?6v&5x25(NR@h2_@!5Pe)W-hA8rgDf+9 zdVhb=;_Q91;pgTf!IEs8lS8?<=9v4hO>QnOrEvDd1p@h5A7bZ6+JV) z&a|t^_UW1JnO<17-?Rt(Hm1A(crbepsb~yGZ-f*^@90Cvbk=s@I#gyDeLp*>frh4Q?+BGL`W9JtjS z-c|~+5v5W`FbH@nlkm0+Q;zZEiN(D@0*vuF)GU6f6dzIZy6Y}X8x@>x+D9#k8TcjfGTxt%Wq&8wR=PWa>xENc+M7 zwLbAmYu2c$jpfuB)7$QY?_`bmsr8-FzmsmKJYqxiPp9+VGq=V6zFJC@1 zJkpf^`Q9(QApi3PR3ySf>t^0Ze%JCpzxR=k8?ODI)7v>@!tRw)Xakmtxba0euHYE) z3u0Vl*Dw;&lBgQ?y!j6GA%qw>k$7EyN@aXABCs^BT+_>i=u`|-2kO&Jq=BY?iV^6F z1ihm4g>E!KgTX_CLdcURPc(5U@$+_A_jjN8$6)9n8U*=-o32U7I$h^-s*1o4G9@}R zFD#;`Ja?$eP`AG3Hd+P5$UN_|2zA(9=o4y<=a zTHare>q2-9?sGjpvwu=;qdoj`#N9J~Gcue0LQnd~C9vgj`xjlP1G;eMFOEz?5 z4mXqSHgK`XcAGHZU&9rj7HNn{ffzkNc}H$tpqC*CZ9L<$(sJ$Sw$Zp7r?`~0J1NSv9ZfXaP5CpnC zgf=l^GDXclz{iqW+C*V?X*b5=`ZFzTCsj>Oq0`j6+cUb-OsvRya6>|gfYhK}+9{Hn z-h!jcql`W#d(s|l!jSRN2N)xm(umeOul7k>l@Dm2EVT1#sXyQ&L99oD52gI?a7e9zyXZkvRK z+FwK>GlF^ig9i^ZuKAnr_X+i>C9EZwB+)XNr?_3?!n8zUltH9%!KrSJTIhl7d8t1q zVQPd&?DEJRf+*&G2eF%h^V=MFoFZ|?<8_d{WMK0ke+WhNv5<*Ger4))C9rTjMT9}*UvFtVyl1G1VDO~aEJiZopY3B8>qxCcf%;^}k- z`9F~Z5Etp`mO4mq#JiGQKe&SFZ7UNAIpPrz`T7)*13`jmA_rI})9aO(YH9R2os-7# zLujz!`%meGs^8{>L)0iZG88cD#LwcSN8C23U;!@TM#LfmOu8* zhBe!&eDlY0hb4V(B|m%g(AcJTMS&xIzBq&beQz3jtbzBf82sXpsIB#3YZbnlxXV>U>%TyJ59W*-BDu#%+X| zI@|Y%-ag+Aw}1uV>6Dvw`Fpd2;$B&!{~Cm2Q;)f9 zDDgBIm&PuE_aW#Fs0+9Xrb-CpDA6zwz^Vj$5tnQt34{jksN#_Zg@AMr2FnPwCL?$x zu{+WXr|ecx8gKTzy{tB3;(>0cKV_G!b}>hV3c5AzDq1JfP&z{zqS~_A(t9PGnhMKrM@kD* zVcGa-$e)lRbQmOXHCb4y6eSVgX%uU!>7?dRkh_u)o7&8~u&<(>8iOMq*d#gzCi&zE z?x4jMcH+2Pk6`L0tRc=D3yY1fg^A3XKj@42igs&&9}`1bv09WqzsGM z6T$RdK+iU-7pi__v@cHW&kp8&D{9~KpP2~6SQAR^e;mdJSNX}MtghK zSMhr}o3>HiQ-d!vryqO?OktbJOEvQ{<4>5_=%uigZ7TQG;>*nH2VVkt+NSbS%^Vs3 z{`vkWMbd`JoBqorEG~`DZ;pTdJ`eXt$em!DqxSpAU~hY@5A*X_M_)t1`;-2({Vp=* zzW>r%g8laxDfn`J(B`Sk*>c~C*R=h%RRz_gJ^ zG!=|)u5~CzHnPb)Z|Q9amQ71Jds!4C}XWT;33WD|OV z5OWB1u>vPxQVUORKYkmi-K7?m{eL{z>|+3Vq|E0Irfn>kZ6H?`InwLC1#kq3WQ@ea zVXH;T1^jOZ&x9ZpX;ogzx{jeA_=`=Uji?BbS(Vx!c81mg_}{44ZvPXntHhon{uqT zA-{U}q+|Qt!S90YcLBe3$*1kDAq&MflH7n`7eh2=7vjNX1*#@px@QQdzIpwdJ4a9X zIiTr~&At_{?=->i1$h0)|8A1%r+PbwgSb~-U)I&2rVA7N|17Pbac@!U78n8Jh5+?Q zyn?rI*rO2^HO5_cVkXDm3JCk;NuQU6AxY8%QSXL09fAF#QQedgxa;AFm@->F##%AQ z!*{xK1XthA@XuFU?>$7$W`!o@>BIRM>4A(GWDghb5O-*}L?OMUo}bN1{%qZcR0FU; zXdkfDWNr@^@Q)(0gT2ARX9<#U#Hs)TpN2;?+`5Q4kCs;fMd7}UI#Bw;0uM4^QrYSP zg7vL9p9v7R6Sf>zHo5fIf(*`mjyDLZ-8b`VfVa0U*j9}+WcZKW za?yQeAE=F$IDBm`1w}!_J564{(5{;Nu{dwW>KOU+j!|Am%Di#>K-$*rsuDM@(6@pv zFsHV#+NSXN@7(9qEoZ$1MljlMZMPH@;|!IRsFu3E+P|HFK;Y*WL7tmr8ac* za%Ab(zJ%~nl$o{>8++G_qjk~(ohdHPcf2&0PiZiM@Z&QdLKU>$e(TIl0wFN;ks*&^ z`iQVqczp8Y)N}53vFC2BENi@d%Gh6f?XOLW^;>ah=DyOC`rCl1Lpt(S#O!aG!1w~h zZ1#GxzCo?=>3!lY@qDMfym6 zbIFZ?QPStjFSQKTAYZPXBg=I&2C+@)8XAVwV=8?abw#;RvXY|i}?+c#MH&JolHf4c_!eAKRnGL3?Cek1#GHy7jTq(D@kgu5Yyx*Nw&)$&psIWg^2NYkv@8`}vV+ zl1PH<30}E4qiCNx%=h?(aseAk(*_YnchYoleJl># z0|*A%vqz=4N5f0hr(UMs9Nsy7xj4*d#rANTjp{ z1hDNqgo&a4RgZ>+uknj_fG|eV?5o61^SyV@T*hzMJ-(+`p>es4c|p>9rkgL?WwaR0 zbTQ(dxz8cap01yH_N`56_^oG;!ZV)uooXYwrdea{o^RFBd-iDd=xE~!LD3r41A09^ zF+*w7*qppf`hM_Oi;adGn89<5K z9*CVoy3$VKTw9j_6x7X$}eDimC|W3hjbj z--56*edN;vwTLh;@{=c45ViEZjm$RmEN#e(m8Qe^4+ZOrdi0vqd?UOIXjKfsJVRBdu{(g!Ua?*^*szwb`K`%d- zVB0&CE|_OWKc2#y;-U|-aP-XgSS#tdv1C9m$=+Bpq?cr#7fH}jd9HZRKL7sYWP3qE zQ)siXi81irW4!n5zh|G((yaZhK_kWu`VDor7c3~>63A0ZwSN6oYVG2{M8LBc8bC=L5SLP{yWA*MT6!=g{~;sAA-UBuX1Y zxntyKe1wdSxOar7Q*Nupo+9!Z7#%uJ;f+JXFM%2+J}|^CRwA((d1XD9CXXE_c88j+ zrR(_!8NH3N(SC!t`5(+ZI7Xm;T%WE%mzlp!2)Hy~p|RgRXzW%L>2DUyjWh;BDbnmc zuCQ(9_uA}@3wS`Z-;u)qqx)RXFaO?;{ph3*O})WDf5{Fz_tMJ z@XL`tZl>66K=^VL`~L%qeJ|ajc0w=o6KW9X`CO@@De6F%@0CZ2S9_*KI3nVN_qJMY zMI+Cfek@5uby-A8#nuHp;el+P!CatiJ{yAb6Y*js^pS}kqfn{sGIa+(;TQ8Bf}CkT zz68%bu-GJFEixw12rbka5{QjI;gKc~(p*v&bk=k}l!5*D)2F^O2Pjyu0aJKuUthv# zpNNT%_7N)-FiDc9R_n!P?_hL~IA|mngW1_*yy%l4?a$`jF$DgPPv|mS7>jm?h9G|x zWg?A+Sep1e3UJ0poO`S(50gkL2tPH4p(TbS#l{#BroLa!8lP5>h#UJN*&X|~pXz%; zk^BY8L4g1ahdYG{A^@SVqW(?K9%1E?s#543$dnJ|nVNK)41+9~mOJ=@$nq})QBa;? z9wdIx<^pb_(lZSTtPB+xsYWGgaCGL-0m>?2?G#d5yger7BdBCXa@gMkdy#Xhq&d~0 z)zIS|7D+8?q{2os^qxW(yyS6C|%c|D)iBoz^Rxy>_f1Lg+ho;_Fz zlF%0j8=-3zbjGAcjt$Xig!OY~l`uJ)cd%ZB&DqyhJb3ty7Nfx6fHdAjzbRX<@W2(A z;uRivg$MrsfCo$hMYDtXGcU*^y@`r1YzOmus!3XZUT^3Ac_fRty8JFZ(g6_aQ<$PW zC1$+ub0f#Q?-PE|{@Jut!RLqg`~&a4Pb@%Aknr>GS~6xuSe7|{LRs)$;N(bh(DuyHK2$ zR~OXqBB!_pTiKOpSD#jESuf-Bxkethi|A<(CWBn6@`|*K;(Ehjb?6D4&lcVvUP;R4 z*-GHoPdWMNn(vY@!;cMal{HuI+Yg6kEY|ZM%(c0E{u@^6LI;0?>$sXT9YrtIZ_51X z0q|hA{>a*`xAkFDhWMDXgHgbkmvv5zQ)fS3XZx)aSkOWEsk!D;SSBDu$PhX|;>0V6 zsy&E^9wnF_J(55IVZ%T}3Oh=l5-J$KW@a=fvw-Z__ken*3z`aEP;@^}Vd2&MfO@s@ ztshty0I%%jvqg0`fA7}T0Y%h=Kr=p$Qas{AY>G{ zJqX!aw?Jt84-iq_fFd0cj{o4fcQ(w!=6R$PMg|VV{~XN)*l-b!YH%I+X}Vv)%L~;t z7xu-fW1s)pT{eTJQ2Zy>PW)B#xLv=8zmG-3LABNVIkV>VEBY%*xbhKGQJ81-c;rS! zfO6P-N%Ou6CKB>PNGM5`XCgApbG;FcF_6B-A40Ck5$vz0D0%V(S+NnGUV-+VMnFes zByI@y6RwRAe&X_-P=TkGC=FNuJBjf50^+$oc~a1%k_AmF37!2JURKq`0)d|$4cs!= zXT5)X1f$(E5kcnureHh!_t`4I;Fk};U_VO47p`78uZ6r}4Qf zz}cA2mEWytU0DR3|1bABy^qinW_7InJ~G9dsRtYQ{E|MzLyp3(`XN`6j?Qc~Z;Q_2 zVMO%UQr~Q1Z>p2|^p-G;O$ni5qcZl~taT#8L@msno3nxa3slAefBO403d^`#?~yL} z>_OAlpMIS#_+#tz8$+}6R#(xCwrZc$yS?SE+iKo+*Jnq!Ns~yEyc<*iN`ki{6e{AM zf+^AO9isl%!@!=IcHszF4l;HCMuZrNQ{!hU$R9%MKjFU?C2|-ef)qTu$%m-`$qCl)3`R~ehtT+x9|Ocf*g^938T^*Q_eP?yR4L5|c;>{dLnFxYIgDZ< z^voUj795$mxtHI9gTs;h(L2N1hb>rp_inIet9e^k;|*BrL1XttM;PFtQRQn;{u-a& zn<6ea$-}6fS8`Ml&|C8*GgL{3zM`7G(y%Bn;pZFFn*|f!pboLmn>+9GN3DHU@3zmj znz!BOL|8L0A$f^;|)s45X4WY8Iy@Ow&tqO=DLjk78&B3<6phYR#E!n zlh!`{%H5`Et7%R1+I1cO&T}#mOY&qBDgDC)4hdZdD0_P<^?;|+$coS>rdq&}IiSS3 z4e6aZ5Y`vo9#fEvqN53N5(kMFAIX*&oRc=f>KwWa7o7ZDjL#i9cZV2%=rMfo!JQ!K zJA4jD!f=ELQWA14%n;J4&-c$ys=`dbbv@mplRQ6yJpsHYxG4;1^om#Z5p*MJLon!J zACtCe&K!Ho=3(}}$b@aD{>AD8UViHUUPhMG{*B}7X@KPCk9{V#Nw#5K*~FXU-#O^` zYMcT|zQd5@U$up%(EU>uG#gW+Sm*nUt-4~5ZG*nC=&@yc@T@gLvEj+g*{#+?mmt?gEakjrNQ$kSI+#Xffx~*8FBEq(tszqU^u>dkS7TtCM(W!w{A%<}@oV9JLE0MX<$bq&)0$m6 z?P|Uoj+23S?^MS`adaEl``m_Ox@BXZZ1>P}ch!+HU%U}b*f!}Fg^#WKDbTK=YrOoV zdHju4wY@RdY_HgxPHF)bC+FRY{=)d2@+SlD3wyE8+!m)W@9rkkbo-zBqu$Opk zt^qx*KY6FA>*D&W;O*@X@&+`=effcQb7YxsCdX_Ga+DnVF5rc*jk(hZG6yFRp%+r6 zfmEQPRH~?+3Zps94-Cfw#LzvpWAN>J@6h~e_II229wRO<_1yvGLx>p&cKb+OL*n;3 z_6Rv5uGv#ZZQnxxT>(Lw4Ek>G>ju^^BtN+WPVUG}XV8MT9TWkG8N|rC0UM{e9}%*T zbV)&!nAwD0cyTOnewjwF>HpTjLD*M`egCcYz&LWZENwzGJ&FJR!2oVWzI>NOz8Djr z-}Z;wfkI#&+J3C)z#(Um%B~ow%GMI?{hMrJ$evr zbGo9yjZ?cNxaxHGElcn3>L1+aLj0KFw!F{M9$URgv{>vm(rOpB!#y|eV7x$T`Tj7vDZQ+oQ==CWM z`M5qh8-$s!uy96MgACsr<>lwDn391aul#Oq9z}{@xVH)2ysx%(c->rw>sL^>A+go` z6CKdi?)4^d^@=WQaGWZfG~^X{jG58`T+urD{n#QVA8s!M{iePFUT~o@cW&Oc%P&wF z@DMoIs(pC)Fh+0eJ$>Ha=N~z!eEqWPWat|f;2ND%!2Otp^=&gU-rP9ME5TAfaaVrd zpHB7OrgJmc+E&4PH?XzUye-)NBa+a;ih^X9RUxnQ9t4{bE(?aA6PaBIM7H-;!TlbU z=?M(}wM2Zvg2?!8f_W;au@A%$sL2jV6{!;)b!2_$en>1J9->i@v{eM9>GJT9r)FDJ zHAC7!j8g?RUBtG*OwF_h$m4(4hJFuYv@0*N?Vq=xcwz1zj_{XXW}R)Y|8ni%#2BMG zj_5u6e8_Ef)W8ETQg#2F{mb{c;0BT~ySOyo+_(r=qJ_Jmji0^G>73s7VAx>g_H}Mt zlGokFHsi@$QjOEO&uM;y9BQkBOYQfK!}~DMfH6S&yg$TWyU!s$x-Ggf$8dS;Dn}-I z^Oa#kIWv8tx3rtI)wHgRYdbVl0ha zZe>Ulh#4umx*YshWTh<#KL^q_ZiL-P;omm}mesDU>` z`U4LWt&q)L4#}Y4uV$%#iub%kwu5u;O#?K4|KJ$yU8R>VBp>)k2Tk6JQv8B-#~MST zl;TI}rq~uie{r{=-&Pv%+JhJ1&aN8z&;PRlxE4zLXJo)fw*FfDM%D4QO*)3$F8}i= z4mxbTq4Op_P_Pdt>D7o66wwht`Ax{`-Uxn+J>s@VLTOO*AR;25`go#Hm-rKyl9Oot zeN7TXvk~dh_<2NQxSXltf74B?cn8L7?%N8B!b)Brz{?sF@)(1Co$thEM}&0NvB*2D=-B0Vl96 zT9zem%aOUX9r+nEYsZ#fpJlJ@__0p($q!}i*$N)(9ey%wj=Aatwi44zv{ei zpYD6R?;UYuLrys8d(W*?Ri{p!I(6z)HGmZk%~5o%0JTJw9)?;I1#cy<)#lFA+gKgaE6c%)!UZ4AoYlP7% zRi&*cLr0;qJZDCvKYAI2|M7P3$Qfp8d6=jhr83AD-n z#IYs%=VZLPm{Z6)=Q0?cF3m-JtSWBlQS`}mO0n{EFrnGtM!4BLe9M{TtdT8vkVelI z8cm#nqUz)-O#f<)>f2xuJdaXT2m?cqErQsHD#^Xf5CAM9nk>g*TbjwE6v76sQ?x97 zLELB{(H5}6Ee704c8vOrcdLOqs%%#}6k>pgdo-W~P{iC(iBq_#N=UZYtZ`BD5#5VQ zA0p0*M^x2Sn``1GAFPszIA8iX&!^@dg@W zDbbc!x`{*n4 z0(CTJ%q$#YZr3o{J}AJtmTSL^L>kx;Ju}^JuEfiEBvO%!&PthdKSo(2dH5RqQekT-H3B zCTkuf1FKJi+2U3rFnu!+Y`#$$?G`fTnFq2fw~hf3cj7kE9DwUPd#%IvQWw7cT3tswA*(a-+XK-x3EtXacc$iF$MIfs@ z2DxLjsML>qggUY0_xHKYRE+5Rd(9|b?|!x?hRCp>PRsjEdJAl^#CqssienJoZW%>p z=d4KX@KzJl>d*k)ey%4nBgWn5`3kFruG>dG(Xg24^L(|QIj+1R+L<7n88~pz_xkwI zR^~~1enX^Q(P@$l2gy%_%q;U|EAis-G~rw0oi1FA-D{z-b2aV_k}@SE`nMJfmh(Ri z0g1K095W6Og#EZ)&=w%TQiJ~Inh9>SzZZy|75SP_l%(eLxM>DwUAkNlEr;BvC zDE%id+QXv>e*q_*z+F~4X0LAGW<30ORI8PKlM8eOr~JkN_CVjHS%Q`T2&C%TA$MQn zNLAOfu$x0)#uB9oBTA}g7NyxGHyW|$zzBL(qomDf+Zx2 z6FMuOnW1w5Xe3@M=!CubdZ1zC=5^{HBl^D3RPafvr_GI;gGSQKGY{NoN!e=@qmh*I zPrSi8BG{KXcqBWM!zoN*e&XK3{NxP&Gd1(vrK$0~+xT-Y{v4YWfA8bJw?{L|9_a65 zO&%K$C2*@-_+D!l&cN!0^_M?}zc$tuiuIRIQtdN3Mg4J5f}FnBd{!r;Gh|x9d3Pc> z57NI{xImc}@E^0ByBE!L6?d_4pbn#sq(kXSxYpEj!JZ+0rWN&&VwF5U6WkJw8hPY- zfa5_}nM3Q-6Qh%(Q=@xF_u*Xa=tdojf1Sg)6`o{KAex1fQIH8xL}FoscRn~qDX0on zI_t#TMp*zd5=tvbmLcL^$jO1&m-Vx__lQF}K$9<=V?CVp8N*?p>bn;nSre|m?;+S* zp$X%f3kP>$kmZe`8PgpL&x@fiW0?!P|5S^M>Y2@j+r5KrdPE8o6<29X17)>UaMZ%% zP|&&ol0ef*O&k5&oe3|eNzARTY+6)=-$bH+qmpZdzj@#i@XymEpo^QRJGXaqT*2w3pzl+D)4RlqU2&?TsrT|}rXayHjC5GYIWQV7aIQd?O1xH~32 zTxf`HGv=DynMJfjT+t$xMqMf{7f?l2Kr@ z8kbMTWeT+M^-^j4^uz=E@836h|GtU)CnoN@fAWDxE|d!I-g~&Ty0IYRzj$v#o)&E5 znkXvQ5FKM1GE|-3JAwZp^4LOewN|0}3gQZ-#W^mafTMT3q#FzMQWF~-%6q**_ajzN z*bZr2mqk1nY80Zf19OAQ6)wWqqEIUOg3L5*tMLDaruHGYnKFM_y}(W@G)2}i*j(__ zu#1MGutL@u;@_4ihPuHiE=tWUf)_|zO^_vz8Za8jOjbnB3;^+Go@E0RQm|P&buWGlRfrvH+Ed}5; z2Cs%Oc(4c`5W93?rYsg=xK&^i;joel2-0mQ%ES_|IVR8e3ze1PLcJhQ(P02k*gGfd zQg!H@^DFznd~199%od`tw{*tREt#;C;?tEEneKL;msefUH>+X_-baLagA7ENIPPsR zGLFbR5vC?c;0n-G`O3Qc#d=+X1Ia3wo-dTq(JXt!09@S+Xv@kV@~xV5t18>7m10Z; z;Rjn;6l+EVJ?9S_DoLN5$lyTvYs4aBb%f?HH`$4<@0$hdtGL-qRN#ylmHkjUMb@Jr z&l?MmQIT6hSFHtehw!fV(a!?EP)5*1LZEs$@E z!g=6eQr6FU4abp`e{>7?0h>q*62P+r2U%s5{5f#B;SdV3OK5I=geOSE0YgjKO)voI zhW;XKG2E0y!YCMxXcSK4A+Ow5sOS}pD@uwlJ`|Vi`K~{@A!ip|vZvJdVq(FzSuL!? z5G`@iDRjLK6O|4t$&aY-9|B`SK}OAS=mDKsYAoT0)HRNsAOs+@M77A_aU+vV+G;q? z6ztI&mW7DI4!-jWWvMmgu2>Wy=5M;k^_Dy4Sl0yhcZR-0vV%=sv@Q5%2rDgrgNEl(TrnBdKROUFrH^{cjl_q5u98Olnvud zJWp#s4mIcq#&Csi!PEMLECBE`U^SuD#R*_DTj*Wlb8AAYC+Kh`BxH{MLq-Ta5Na~n zk6ccyhd{+rh*J+}+7&9S5clwU8bP7%zZI*whumnXwh;6V&ijs1v=Yv%wp4u_RNf@B zf}Aecufc-?V}MUAo2z7ufTqP7k~dKC8a^+Rj;W%Gxz|hk+d6@^hO%=TQ2Vjah`57$ zgG5k`tp2Up$^!mU2uQ49-EW3fA;H5s(K9nNGOgaf^{^m%z%A6xfIDhz?(1R6QqG2z zQuPwbtgci?wlnY`-ZL$9(*)|V&<%-6?AV{1ls~5E$JD*-ZIh2!pV4Ff!>F=}vu6Qr z(gSYN1~=&kHyH*u$>63u;HGSFQ+{w$VQ^Cb$5aIOi0rQc!)F>ieeiR;AMwD;3Gs^m zYq-Hb^VBJhvo*V=(C!Sy6)+Tfin6}=QRs_=3*x~K>P5j!L2E&L%i$u5aLD~*Z~-wF zwYsh`>wl?nEdMg;vw`w2_M+DKlGUki_F&~GhSpwa1w>>tw$|_uOc_e{qK+1Vt@$Zc z5ltO1JU45smcG8#WHIYMmz?zrKj9(LV*Rza_M;XI{##>E{(@la@AS;r`hW2B=`!{) zhq0_)nYmKoj!K1#j5@4&HKWB(Rv{=qXPBKDJWW0iz#k_cfET~wVMe?G__7#)zv+1Z zUfT19b_3KMfM?tRkY=G617#yljB0ZcwGs5+W-wI4_VKt};N9W5emJpedgTcZtE}O< zU0sWF!qe4^>F`!#S$-=pgmSu_{<(VfTD58HYICu2`tTa&>(;*Tk{&g*~0a~?&k%i@bk^#%U$eu3xi7gOpB{Neoq&*c|^ z_kY3+(OB_HsDDY@)W+^58@luOD>&`vN zC+BDWiDwd6lk+a=maP{z5fKSLG&aq8+l8}fHBBSiY0S&}#f074^Mrl%S3I~n6Snn) zqjTAsu+59PZTlCmblXlPNTq?d^Ga@01*w_9!#9(B_+I-X4@<1!J3ouPGiOT-g_pwh zF6+k@f-}!R1JmUq{8vU_n^@gqHNROpUs^|`Zh=d2)30&si+Q03bH;lz6sJ>bLmsa^ zbo~25B4gFQ-fdwdX;7l+d!+}W<$H^+J`n1v^KydN#t9YM7?)W(dhv_5+^oz9MI3MX zeL>`^fm-3B_vIFJ-=<*h1eUjFMX+*h;X(lch3ii_rVQyx;2V`mr4O=RrCpk#87VLbwnm3*qRQEF3O{sVx>sAq50ua=o3_a z�gsa?CZFh>8|pkiU|hL81e82@RcHtTirW%jC_>A?+i4NkeBFa4TfX8SzNeAVz47 zjs-%wAs~hS;uc+Pg=+??b0wm{aT;ZStIr=rhf@Gj4P<&YbvM>%!)A>|j*gfPr@>Zh zilhS%a@`EcwJ%`V(`G2n2wV8ZYwy9w@wb(4%fP6U7yB+6@T@&Q5~8H@47EaXzGUW` zqLSy^+r_D&M>I{ZtuBo`lyv8;x^#n9w|j-(>bFA!PoEC=TGz9`U#UU)OKU!6;t!6=8A~WQZNt2lCNU>NHh~F{LwpEcq+j;7u{p)F>{ctkn zXAztfl805Xr-Y@0EsX4^9_X&{YuCKgANse;{NGQL`B|O-an=cq1QHHP3p;vkY1iJ- zCV-{zj0xb_%yZWvdRT30eXe=hx=xUpEFJWhCMMM{c<3)pOyL*hWOiNjmnLY3V|>zK z4?Vt|bf8W;P$xS;ophmw9rb`Z-e*O+sW8Dy`k^&<)SM zZ?WpVIgGHA*6>=Or;HB%+okp$p^szP%)8I(QB$3>&HPAGK4{!pi9vZ7({*hu4`Y~- z7sjGZNdnQ%7#mWsd^^x3id(F zw1n&O;5TXwZ>EK{F3->j9bJJ#p30z7Yq0B1xxtW`06Q3UK2TUen0nChm>>m6n)-5` zbPUS(_M&Q7HBWuYL$9L-QuX@F9cY?o6PgJu0$-P01P=bThb5K>e0~hv`^TCsJOvHhfBGklP&GadYNr|eH<;vKU&oi+7^2xA$`{Jlo%9b z`ddRX?XJkTJEvGGut3>;9p1Vk^PQm%Vy(}o{#FZ#LXg+zJ9UD)wqdmR(ydh_$-5nr zSsWC-zE>`>pGuS18BXq!L<%jQqyQ|EbuuNhzLchayG;7!G?{dxo?yC@mU-quIofu! zgwMN3O~WSU4WlrrDW$18&2*Is+0wWq#qX8sGN$!F7ckbqb!`UvJ|E(yJAwN$ssBp| zWUMJWYo@G1iKpyt4-3pt`QCbniXJqpdoyUxO1tBJgv-)(&yiEp-ALd8;vZ9YKX&Sf z(6K$8WZz$Y_v5z@UWL6E{wM6E9l_WU48A5{v@RRFe+xC3aKE!~*DWgQ!e}L9?9SC7 zhcn0!N5!Kl41(hy_)abI!ENCIrvpK%kuAh3tchQ=+BgA>K_FURo&t!?Jf{&h-)djxob+-1UCM zwq6m0)(eAwESDC4xj=$QU!mW680In~ z!bhD=6aC|9&L4~Lbm4kf)?LPw7aFZ@b?3B&WB84y7k6M{i`=@-Q{Brau9O~5bC{2$ zqMj_?r35gFS1YhZsVaN+V;0pig%UWgH`sJxQ2TMc`WKRq*TSN zB#uBVHcSL6B?J>0v>N#ptacj<`BH8C1QJN#(&uQt~z&XYVn;T5t{tDUE%U7QkHschQ{R~StE;hm?8Yo#%y@G9XXvQR-dY@;#1 zj1uNZv5OS7$3+&lIvQa$#AIz0D^wF3HU*rio3iZ25=vz(Y{9o$J5S4_#A*@Qd(Fyt zgJhN^ruq>Es1_0zJ2Ad4)l6?E6)4Z0h5vqi8Rf#Y=I4 z7q55Fn$>BbEjD&<*PXX-YLgemTF)E9S&te(LxI<~ktsyoQaGqg^ z)u^%Y^aBLAkSSz1qTr|>eXe`%mF>wqq3lT(HZni(#vL4O-;0;$@wI?`KTfPB7UR%p5)T*wd$u%pW^!Jsm&FfR57FgPFrK zvorENEgAfDd@z$KEpsG%t&v|ZG*|P*i={@hkqvj2JCLCxSF~C?k5uOE?>hm~h*=N{ zSddEfM%KLR^tD9V^R0T5$zab11|xC#JQ0uu_vTAW@{4-0Nzn$1xH9rcnO;YU)6`bL zAXW>a$6{Oi z6pM0(KtRBzA<%hjotNg9HAeqx5rhp=yf_a!rjHgXjUu?SCpdDkiQ-8$@+mB-n-8S&sC*!D z+~XlEli@wXco2vS9(?de?v6!1T;qqML8As>G*OSJ%7fA!uvx1Pb9OSB&NUAmQ>@cP z5u_rk)L`t8qkt=g;d5jwlm3Fd9FQ}^3gYlNF*wcIyo8k%V`eu&Mt(YZY)lU?ayO6x zJ+R`BJoQyAWOG3HP_2qgBu$YS_k8ggLh3~E8E@a?)4?R_2@umBtJgNxncBmoHR>36 zViVV1g!U93kkMd9!2#Z~?-lxTVJi#Kp#G1NSx8)iJA}{G>2GcW#v9n$`?6NJSX$dy zbE`XUwRb}W8@95Q!rH^lkH4^~#n<09GU`p)JubNIHpo`0cOuQjsS3g-Gu zZghVxmzg_rV041yV3Plwq(5i)&nfLs;6U<_Tzo_V;YEAy4zDN zK`htd(rS`*E-K!MbuM$5i)Wn>wX;}7=+Vi}<($`fo@R`f{S~TQkb8_2%$eaytsQ2b z%cAsCj(ac9BaS+L2r*;%l67VMn`+h;87EZA((o`FHS zvtTdfU%v%gZ}a(t=9jR??{L!Od=}1H_mEL!TxdW3QnltL>7WtGiLJV&uo6RunZ866*d($M z?c{TW?}2@B6fu82BV+meXhJ=TV3Ywq3IT*WqXg#)(5p1OjRZ{$AH zZp&^>G`{7VM_BP>=%3!QE4|ho!;kZXq|L6#E`fP26P1U;D6k)w;S1cDa))CF%(eGI z%0-W>iSHAE!c`S1j3RpNV~ikVqbClhftICa(t(GG@wU(ixLWEChb+wqrHwkh*5k}_ zd3UIbSX$?2%s49)S?k=X0}R{7z+z9gR*^*e0)J(!z3-Jv?7lRK?QlU!F(iJ*7K@u^ zUSHA5>;(9jP?>$^rdzv)W4!|h#pEG`|FKqwUz7cY%5A11{B`6x$2HMabwv4F$z&R} z$^{j9Bs5N_3>Ngt0{pZb<|P$QTa5%!n;m6U5sw9BDsDi)jfIdXIG7QYzFCA5iS(3} zL1rk(ib4sdAk-?H*UK9fWZR*(#fyc-ru-tOgQhe=NHnAUvO`wNZ17&xIzBtBiDjxtxspw>txuLqKAwn^z~atSBT{-MskvkF~|nD*>CxVY(K{ ztknX+f4jr8mFDogH^n(%@i|qnkqmo-HTe|doM$crTWK=zMm^beCoS{LgL<^#Y)z$3 zYt6sQ9LhSD1z)_$O`lz^7Zxu~ZeOZyi(&&P<)kuC(j~DhL~G3AmZ)8jdv414l3Jd< zbe2D`d)e9^vwvzh+X~EV@85_|&u$xo{_iDiS~%FC`$0%x!nOIeXSKdHLSk(+X0h=W z$0W3p!7Q(rF1kiGUtJfgH`OD@s!4%BwBm`@WAcU|Lal{SZ&~AS@j8pB!-JWuKWn0!iK^FvxD&A?hy`f*OCqVglsrQRAc zRG=xJc-PVI7#2AE;f-HQhOTG1E_h?N)IL`#~=4HwMBA|-|Ei8 z@nc-m5iu5gV}k;!G|w`mDHW4~p+k=UVW7*0mg4K$3>5u!h$3G&xu?x74r*3^p%eU)D z#>i!X?ehI@@I9e1F!-hJQcs2e;wMl%3F2PhNM;r&1hqBc#6xXDtni5D+brQ=0KUj| z_#|mwP$T6f(%W9OkH@jMU$5X&B!8hE2}wI~F59^DpZvPFFRyGZn}o>HjeFBk0-X`+3L zD+s!$R_KKX;b;TM;+0Ve);f9ftfj{lz3xMePCz2O>(JuE(lA>SBlob2DaRA zcrH^|9mlD|Cn@f>Ah46}fFe$tKA@vj;cLtX83b8ZxdZZyW*aD>N=*=RwzvUn4@S#E z5d~p!UxpMme=VaAE2Zq9 zPkO^(qvGe@fy~P=mf;c`@-P_Myj^9#w^&?bRxBX|pAw+MzJ|B1n8S)uFBNF} zF8~rW`HoETLPmz;<5mS~L&z%@1fT%79w-bbfQVXJZ`q4Q6rxN-t%|d-LN(*Y@5nW; zS>KQcYWYBHj{%a3mBST*)nt+ViPaLQMFc*8h{R*Kr$u1`MI=N=4g^}}bWICnZXjPH z8sNgF%u}YC2Xw}Vg>Q?u>1c;bPIcT)C?aJLc{SAPh~9AZw1VOlKuVnXS?eYb>Si&+ zD(SIF^VHxgf>tyQ3xuRQGVEpiGgYms|<$0=l?(xs87XLleXtct%$lhK39 z+Oy;=G~+F*;kLJ7OJ)U^oma@6LJ?sChz$MGz6dhtx?*ODM|4{- zuM!$|S7vm4RQGtPjS!Gnrr0H8iYX-66ni)Mi6VJw(HZlDNMdjI4wJG-JG|AB%Qopp z(`3ZGPJ*il0nvE0diu=6l<4VWjhvZ2?(Y2GXD*$U7epJ)Eo{!tVs6Y|a)M_#Vh1-a z&=k9&)diV(Zhd=u+m7P7@{Z~MeiEl8h$Z<4Ajt{c$S?k`-fI~#F+PUYz0hUrh}IH& zT{wc*P$)&+uu(dojy$-fdQr{7p1#o(m4BM%JcMrD%_^Rup1i{Wcevo zmAtpV(72Uz4NwQYn7O|2X zkeRCFdgS)+4Ly4JSv3E1a?$+E3m##xMDv|+m6kXfkRvEnOeE#LfswD|G@_Npx)sq7 zO8WolS)O100S^hTJb$uxc{bPYSdng%zMdeR?Oe=zPDcDXk3(U?U19)D?B5lPp&mV! zxvwObxz~QfGhmj?-TEw5N2|i`%r|P4kcA&>yvnzZ@%nx<{nOK3z%TxV2a~T2(J_ni zl`1;GP!AD{&Yc*aghu5xZ+%m{L1u~0HAW}VpSu6@s1T8no<(H$!=4^p5t(&v@y=~S zDOgHki!D9tLAZfWKbXd|(EP9yVD6^E4l-iO>*#Fuuz*3eLwVV~KNM ziSS;T<(Zv($X7&XUaT4@NMm<9cr*Ckw^(W(4+(?o{CJl=i0xyhv{FHb@>ntRw1?GdmL?6gfPI&(xUsR2&!=}t~}MyvhECEA;GnMPaN zW;OjP^`?2tv#6YW79IR)J<^7eW){ueeC<_NVR5A7|2Snv2salb715>SZ*_B(Vg88w zqKJj^CDjjkUR2J1PO49{4Bkh|d>p%mzH5*Qjq4n=kwbDN>{>HR|-kR@)%q9r~<1OO#w zF@4DftEs?R2f;JYnUw_tQomc>vnh%*~oqRfmGcUGum z815q4!BZYbV7-NkF&*4mrp)HV}w#q0~T=ip&y?(UcVv z8i)rt5BsNr1UaQ;%5z9(`%G`8w!)zSh>2Li$tB0Z0W!c*MpRXwY)#~!ZXny>#nsXR zTq+j!rDlmJTrDlGX23ASWvB!t8x3USq(BH{`lRUR&?RI=U1MW$m85B8t7$u7Zb59Kw70ofXz9(LW3*%@;z7-CaxCv(a~)dumXESQn>tBs*`AYE!XB z|8~jzku=G?del%+{sa~y50<_l0VPsaSj)|}t~XFy?sE@(cV?kx$9}~D{@bPgvuRSF z?`<`*i0buA582(>s)AZv=G>&0L&YNNiOZ-jq{*nc%eDAgEWLbv^v!$>(M)3Nz#=gP zU`c3sG^w3pk&+x%5=xGwVlY|cqy#VtP1A|YH;a~pfIKvmv&f5+`_RkVFK1V;Z(Vcz zX{CMWPpx*ui-Gd)d=Fl`@-RRnWnd?BWp}$pYW3)Q8;Sorx*4>JvhIxx&E4ApmwVFW z#%M3rw#BF3zVmRfpT#Q5ks@PIqMB)5^Pc?Sp4chx^mY=itMi6O<%dWR+wJ~nXNd*f zx5sb&2W+8eGm9Hk6#8AOj*J`#4%dRM+6MMUxO2rjB*f3It<$A zTUCR5p{+Ec?iZW7p7rW3cYvM=4UnrlZgn}KW?T5N2H6U~!3doT_0WAgB=yTVvmH_x z@^)zZbsY*Ley!IT5%5*NI&@R0|z=qd0)qjbGf;M#qey}Sg_ ziO<#htWm-b^!Q2)UijMVwIQL0c1Y2I5>%rtZAIi}G-Hrb{k(tTNW**PXXAWkL zXP(R)%DmT5^^lFrLxjsO#GuO8#)4G-48gELDSv@aTnsR~Rm$M*HguMM@|_{)DJZ{5!JQ=Kc z8`gX>SO*QP^3zEfcH97KU41+mVo%zzo=gVI$~{!h&;`3BQt@6JG3b(ZSkd~c7@@zQ zd!b`dtS_v4%byau<&kkvVFlm#X0NY}rN-MqDS3^?zn)E9|0=EdbPURTZ9E#&G*guM z*yL}OGB}UUH2rqgOQwxYL04dQ!>|JOgYzh0)38_EJPNpfV3q%_+1L$Cpe4PtV?I9UaKACGPBK(ww;Sz=3>DNUn+3I zA7_xP9C41h=VnbX<5#CM?dIWglhujDq1rE9=)M{pAO|3Pk)@>#gfhCv3}=h$jZ&pn zMG&y49~f|1Xq5p+Z$xf&VOVQe_zf1ci=fF~8eNC=?e@Uw{<$aE=pEj=O#4u%gYG8{ z(mnmk_i8pN1o@=lc0c91$5w|19zvr<-a^87-dvHdy8*mvfI>H7LRVi*hWzgBx`)8O z9$jlb^K|&GxhA7F%S0VlMnV6Kj>k&H6Op5Qh|!IaGY+_ zxf#-O@TC(DUUov0#dQXq7LT&hiyOTvBZx?srx!Z}iylV;xH_HQbE*qUrAmkU2+f8D z;NDa@bx%T5bBOk??nqkeVN^?Kl|Kpzt-C(@3YW=4T;D5qy#68tI@Zc~%}4bJDb#pn zEU9E|m!Hp?iq==LQ_oODuizH5yvxfo$El)ygHtgg|L#Ve^ZwJFkxp?UT+E|ULX-(E zMxeYZOQpEOfusz4)NE=tUM}!@0Lqz3UixmNe=Pq-S4rZf?|LIfUe5fy$lNF|hcue? zLTP2SiE9y?lwOf8c~GDY?tcg^AP?Tfg#|ipr**rw&@3Q;hO~s6Cde53Qb**286`eJLS3nuV^xK}pDAs8ZfyV3aZUV#yEm{>w&zr9duYClL5MLc^M(?zt-j zx~s&fvJN*Ip;4%@?m^^!)c9FEi_T+AStJ71jjKw93u)M_bmT)<^fK*=goI<{{l+qT-5$4E+aN zoa%ZWOF8ax<cyoJ7plo@ zalupoo(mr3j+E(nI-@$Oha{8=rq%2nyk)2;=?cmJfe6lNuyFcPU2 z5PxE(0v%q(8h8O!E=M^ulM)uFI>898M2J-bt0YW5T~Vh{DK)n!&T?er5yoxV0Dm8E1X!y+Zt1s2<~b8!eeDv(tZ$zk9aq0k6rb6apAhVUoA{ex!ktDLz4y4j>lfMtZmvWiG_5ydVYVX7dZFrN^&j{7l-rSqlc*l|jp z04qxFyiw3cB&?{_w}M8cgk%``3@&d`;WubG^ohI~vL`88>$G6hHmd?@ zg@tXhjrKyiWuW!;9oCUQ)dC^Su}2UWX)f5E(nZa#!Jn=Wekt^6*Cj~pei!OmhE&&V zg3TlzJA+*-1zjFG*wVe7hhpEx zT8)}AiqYn+9cF^nb;l;2Ufmr*h;`Ufqeu5)7d{WA%jYgr*&_BG;$jRm1MS>GQ=a0V z!aQ_cQZ&<~`@4P8y{9LFM`5RTpnSzRXE2|W+~1*fJ^U7hS` zY>4OA;4!Y`4zS%N{mqlX7&%Cg3p6Vm;(=-jtsYF}g7Nki_=HY5J*l6<^l6)WaPx@Bbd3uq_ezPclg^|>Z?-=h{iwD zobh>|u0@w>_VeF*JX+$_K>5c)0JD%n-XW2(u>$W(5;n#qWHLeM)-l-I3)Jr+C* ze-k4fo&d3C;peod^f3!RyOX3d%H08` z#aFt&jG})Xde)3*$^8nr9u}HdlKW*nTzyFHtKX}g8J$5d5Aykz7j}F z{IySM=7{cq&`kzM4cC6CQR}Yz--z5b7WlU@QYPwX?uX#Eb?RSIglsSMFGXd}j9ump z5guzOe909hdnZTz=c0i0fH6m&|9u0jQYkK^FnI5Wh zHk59Vaak5&`jH1wOscEHS5Xu^%WrWs9Z|NBN(*AoG$}m|${$HMvn+BFgGxfj(R60G zMNV1}lTdXqxx%q1N&{LFYTngd_gR=mwe(=ole+mxn(Af%u7^%V1}&PpHuDhFT`gHt zc(-1e&6;3L1K(RuZ{4khQCJI>wk2zUR6Vrqm^P%o<(NG7P-iTjWpU_iqqMSCpn`>z z?zxBz(Ui)xu~Fx^#tfIv9;H(f=?A5P-L$&58E8SahGNy`JZEyeI)pQy`WlXNMKu_* z-uA@6$r_?q&B44M4vCK|dp9Z~l^tjv=wV@t`=>%+Vl8@qW)?k#1}}QIdpT+m*P(%j z9G~Q=zSM%~k{q$R|3;dGzQrdW7M=e09x`1Ll(8;->FTFDVqZv;khf_|J|xT*F`e5) z5_B!5gpxr1+a=}C(WK%Nvf9j9Tyf{mZTq^rOIo|IcGU~( zFMkYwZLBR6>o1?Apf(CH{P8zn{meb_#oyN3dn1_EiaNi^Cy2EE`9ku=V2koJ1%xjY zR;Vn5YRaKf)_MyYMFJ^`O9b-H1TB+qYU{|F$)1;()B(uvyxV)TEqD@7rHj1pT;AqF zz+)|p*L+!XQz6F-;{fdjJMhOMvum?Rg3S^y*AMf+CGDX!NqY-fo^7M5J8b#idx(iv zq84|(vbuRJq98PGlTaB@SbnhXBFidEd(WqghXO?Nh;pp3G+C-hhNG)Y8@@!h1&^K% zjg)IH#H>jDTnI=kNxt4lvORnRg6_*8gbBzrc=UpWHgR7icP^TzOX*1SwWAHC9X4QoR3=S)k>O zCHq?-+2O+r`vfJ1Nw7P(L!JE^k+V#K_By~~iGXGe=#C}%p zAx&bRD;-Tejx2q@o?QA){gP+sEb054*bu=|bKB$gPrXs& zJz2JaE#Fy3(9{CO&Qlhp0ZgGk3-6s?B`QgMt(1ZS}ePJ;rF#vDAL&OzC$oNJa87j zI<<&o(%U)>zG6k^1}?#Wl_tTrg(c4-wf$Wrk6Mz__0gSuKliqA<+oFF*H>y zdfHw^lC!&go+*-iuUxX;o+eqoIxtFdm7_j-n`4Q+*X9hDJC_z`__3z5yq9PkC{Ol6 zD%+b&W&#>azJ^v*?K*8;(JLy@)Puz<>LVDbgd3N7=B-^fE*KHJF*@#O*T4$Qga1M||HhfS8L6IBnrohg^U9{O5H7+k}` zsp!*=p3=op&0J1j2z422xq0R5HE$KlY%s(saIX80T@PLL5nr`i2zFAaaVwO!Gg@vV zI8Uf+>$!!w7y77&p0GE~TUKxXIQcC4^bhHgHjK1p(Ot)u$vF#SovN3K>z^Wtim)1$ zP)4@Xn3sQmByynq{k;e#ucDv%Ef21)Vm#tx@OzvTAxfX)EYfjYb{_X$r>um){Q?eBQlXrpDyS|RNK!fLp(xhhQVCe&gN?ohxL?!r3*+zod2YLCnG zu|uo8)f*DGd+5~Sfxb~NOL!6>5U^hk37D&FB)=FBJp37 zGPIDx31QGwTT|BU^XSP9Bs!>58bTzQQg_Y0x)X=jS?E6`slvq}PCAMwxC33hSgbFW z8bzZfPo#>0CC@1X#@x6azmO$u%vyic72>~$jQ_Oz# z1TIl?glX)SuU2=Oxop_npM1jZ{x&_*hLzTYy&0yG@%mrd@6>v#*Gr)uS5*~>ZWl{xC+$&V^Bu=(U=CrYY0Gb^x-|ZQ{-cZS&;CQ9g>Jix|inMbk%yyG@8gJo#%9Ij4|bP0SDM zH&MKpvO%>~6A7ZaHB#D&ZVN&9j*u8T6QHGH-Q^@uKV5Ere+WdZnXoXc`KwUineZkt z6D)pN6nS1ok|6D<31zhF-!2iKPLl|1*>FBXBuc`d!FW(vzM`dQO|cv~=dRm=#L0`EQLu`7cSL2g-lmi(s>< z`Qj^{K3xI&kpu!X$pR#{5xA@Y!689`D6yq!b{<&9qzM`hI*TttW*z(*!dMR>3(;4S z3(-rT^9-mZMDIU}>w$#;In0q-k+bDyJDE*6GBcNI)!}BK$`WvHbYz5*vGI4`9R!)f zMdT~71IC}~y^Ghr;gelDG3vW@vUk|s&?leU&G);0h1F$!^YVs=7V5$H!+L9@f4jqW zdk8}36}+_zK5sa+UtYZ+udJ*V9TcofS~qIO>fT!Bne&y%<`TVJyz-YG;c#dBWajKj zVQr04-M7tBF^{nTDDNK>x!Lsh@uen=rGFf3OsGOX{nk6x2EnLMm%#faf%BBjxC4pp z>y;t1@9$k=A4yK^XC96uc197q1!BWWH>7c$VU_E%Av*;~E(4xS&VW~okqnqt47hah zqBx8Ttvo`evHo-$^Mp`b!sn9{{#q-N@P`cHTgFLmR$M_h{CC~xP6os!yOx~nFaETL zY-?@13G0=p)1%BJ>xssu{2^ch_4)PmPfvA*uK9wee^NNL@8nl^Woh=JFB)Nt27g0GRl~M zmj^hp>D#Zq3U`|EzwMW^tJgEXzBf6)26G;M8PT50qt98Y9>;zl%I=U|DlIP;>r}K# z8gsO1uGXq!@@KucA`jYgL1Tm3HZnC-rm6n zs6AC$TA>tqiqL#;x>y7T)SguUxke8r=bL4KrU4Dg=f^|J2Y9xnO7gUMB;f#Bgd_&l zGZmI?d%IVVEbKZo@bu|=+@yP|q!s}E+tqNt5n_L5i=A$j%cOBrPw6?ThDIC%M$5fL z87hWF%@W#`s2)^ET6Y1lSbHdiDQ2OY3b-Wn+!J%U>j7@X{lbG)^cKnz0<*&~pS5emk?^4;&iIeW%8leiKh3cP*m6SAR^byrxv_|tfI8}_kd9V!q|dK4!^>B_hYhO# zH)0ti@+dpv17;rmPyWPO=YAFI+`%F+#sL@dw%4j4Yomw~FQT@ki2&Tfg955OAdv&Y z8VXdvF&{iD-TpON)wp^aS9Hv%0+_0*CGkeYU2HU{sW{a|9WPoK*9z6Gv4z?a4k!wh zMJ!gdHlfbMEQYQf0I=knOlt+y&n#Rhlu&x6f@fHnQBB5_21A#bB@6-zgpJ~h239Go z#C+(16<44^*?Up7ld;7*F8{>=AQo*pfETr6GiOMMke^vBnj9`tFBTc)3^=z~Y@Exk zHk<2>1LNbHo16Itt<#IBI9JCJN*)XDxO_5B=lYO{@%2(^{Pe`$`zI#%-@o^P2PP&a zCnxU9v;!>re_Bju+f1tVj=< z6-mu&w(NEZ6=twAMm#rBvnW?z})2gQvEz zt)!hR)@jcEgp~|Mm$6>g$yC^+$eOL%hFHxTSe#MumFy5C#Y1c+A`1|zS-?t&vkqE| zOD_n!fcK$G#YKRTh5}h3wB5yeUCWG&tYf2EXkb|^tRQj|H|D#^%WGQe*umx6kniHD`q7%wACR+rh0YDggj*Qn!<$j+5`Oyqd_hK5D$l=r@cjt zZ`#l8Viy^en-hL96osQCr{sU4xeY^|W~EyeZkQ&f@gG4E)GCcksfw7r0_vYE^DHEX zm%NB#YD4Ri@jUDE!O@L6&FeLxF^ZW6&R5UV%pMtGFO}O*7Bd|sc1&eF_F3V;b+rW7 z9tzr3Hc;CxC}dW*){Au=LSXmD}uw7}mo@HMtIQh${5cL`vFVY;n}_Lb3e@sWc2wUYowu6U(I!w9_1W zXLp5UhCHI7hq&mK!9vQc4E>Z;{zaq@-4h(Dt*yaQgodxxn^+C_-=L=dLekOoE%JR- z*VRk0kkhmECG`wT29+GHlos;qTl5#dt~BGmZ=5GzT{SO`b3|z;0)gilc2ubbI50u& zkA_5;(tcOxMP>S>d9E4+qjPGW3wVu+x2UdDt86rj;&le4qMv1x7|RLOFlJ*Hb&Qcb zoURSwECzNcTkMU@Q>UJM;>e-d`BP6mIXi#&*eM*|4W8ml5sksj(PNK2ed@^kvBTEW z@uLjrD4arQad>8SM!r7`p!{@vFq0`QbLrMvBfk#&IA6S2LZS6+xHDO85UtkE!=2Us zz7rsgm<6G5o?K#$ta;bzYl*byTlFRb-`+9?ps-S$CjwCNePzD1=+5)uPp4bX-qhvi6bZm1an zgQ)|VYi{B=uU;=~2?r;OU&`0f&pg0p1#Y8372C4Lbat40 z+;FcBj|RhB4QgvRm!r>69-V+}p*lJPXx8Rsy_~G`c2tzEX-p5}coS|4kR>|BfAdTA z+IqEsBHyae6ks`DBMwerEuSlF&Ve-n+hIjquhNzznWo^6IwK%klZp(~P>qK!eBh${ zdLuK3PEf$@g<`#ia#$6NfP|VG-3Qv{&KwxsOGL@1BO}>;`HAtp`3doE%6i9*C#?r) zJY_eYpm&q33xuUi7}{3ba6=t5OM{sL385V7_91Mo4sucRJfIt5nZ?R7fI?5@k8`}? zDSc9#&E-*PK07xWD1PQLqCz7+z-M_lnU_k7XrYD>hIy#GJu#J(T_Ba6tUP3uf%p6= zi-!%KX#4O$@b3K7@?h|8j|Mf>@pO4H(0N4L-CQ4n%22bGJrGQ?pS)OKrb`Qw6e_6{ z^*@Q-@G}sFf(4H-8wO@dkNfdqnv`yj#f@(GC|o>H1*-@l;{UwrQY=aK#@wfAw%(j){XxSSKG0Y9dP3%rux}tK6Zy%^yz>iv%nOMWg2xcEg$k9Mmo0sw zr#{i2w+kRQks`={LjqnW5#;Wn0In1fH28RgrJIaJyh}p_6>eGtg?LzkC?7cDh;MX<$Eq!$h7MSI~Q>zX8ZB7Z;5LN}+UM!CkPVjuS~Jq&4t z`vFIKxT{O6Yb12msqL_X#p&U-VQ&{u?U7KKga%;C=Ikbvu`UYd(k4P}0o0ov zGk(X6mns-4h&8a*$XIN5%=jHMo�Yy4&erWw-2@@iNmFAE6e49W%ae?~asdx?{#i z?<5tkcFcI~=YoYTHVJ><%y@lTz~>ZiexUocK)Og-{^%`NT~+=V{zLKF7x6C+5^lqX zB8*R+w7AEJ<&UFX!LB1hA2;aZ+!F^kb+Gp>5>EJTVz1@X)!D@Ah{DGTT6rJ25(*2r zc0o}oh-S6&zR)r9Qc-vpdA~WtnJvgSk2sRWVg;VTCcLc83^`M!Yt_V!(7rGZXmdW) ziv0=LJg(U+uA#bzQP5PdVlJzNgQ5iwivzV_wvJ#Ry4FCmdKQJw z*%_#GaBm-J0VschW)b4h@K1Ol3@K^8$RF8}m>vXCenZaeu=6Vf`r^23%#SM|Enfn5T%z3_PB$qsRW+h<6*3m*lj9SWY z@M`FD7&f&rxTY&0U}R*WRA`K;V>7TA=qwd$%gg!Taaz7Min^3Hr1H1VLKE@2U+ySYwzKJ&p1+;SR0o+T1V4ScoxoSK}^{AuE1Iqll$V zqey?dD1JFj6z_?*gc=lDali188G9ZcNoB>hkWL9;5}L&2moD~$g>H0P4<@djzr`^| zlsJW+YQbuSUI6^N4k%BPaDgOJ^}E|`A&Xq!TMvQp#U~Xr%R=x-IslW<^l(C>&kS`^ zNFFN8ZZdqUQQdE|R3vSk#6GZ}aOTAuuDUAq=w&zb7`N#v^Dxj6=SCYS*&FXI=4t;m z&vrM?PIm%~j~%9XT5p`Qia+=V7@0tE)2F7gTk~XijiSaXB|)gTfzy-07=^(ib6Zw6 z!~@k5d&yub7mT;Jz$X-*C_l;NL`#m;K5_bp>QccGyszTipSr>gOf(RisI3 z8p;m74&|c~ULWkU*&W4>cIB!fk`f$F^fYVY)KqIqu^+`W8!IBr^I>nrPV97_4MermImY8&T?DWxzBU>%D&ID zv)m$NLa5}9x0803TYbo}v)tmOX=k|=i}01Q-0JN)dm*2K@1@&;WXJHJ6WVE`P7z$U zP1o?Cgwx36vQ`|V`(Rbk_G&qO@M0f&l%E_tq2(C)dr#x$s z?eqwjw}`m|<;{peCf5n27Oel79+oD5Vx2k{Wqyzlf`K@4rE}sAdX6W7-2TN&b_&goY#t|5wxQt={bEpdcdK=Ap*1Q|)xxi1F7O-C{^ zf=iY-_*|$L>RUMT<&$NuWVCbV@lw}v3N)7lPJ=unQqprAAZ z+^I8q82c3yi}n#iB5Y_8u4zzmpnq|-NJs+3gS+dI4;`k88glMKts~k`UVk*U z08qfEJb7&z)uDLCC$HCqvejHeNo_o0+LiZ~934($dRya=4!lGZrJ-wVic*1>grfWU zRD4Dz`-R}a+E@Xd4h$^OrB$yjA>IJ#VWqdnjoPM_MXkwQ+i1GVuMzx#^6&LB!#HW* zB{RJ&9B*p9vBz01T63AvlXEa#hk=Lb!fJiRtTcOq91w?!;aj+U)a}Fr^e=xOw zVq!w(R)F*rn~II;0W`c$4<@AZ)(6ETP$ooI&a!lVof(if$dAM3ylppF17dv{cR=nx zcJlDkr^91Hl^bx#fB@=XfvhbAL?YmiGVsuaaa)Z0{@$>=ZxG<_Jw0>x8w_{%^u}H5 z%edTq@1e(Mz}!Qtg*q9_r(o z40~|qu_Iv(3`?O{AFC}cZmgHE8g3xbFVggHlq!_Mr$pi2wP5mreYzBGA{d6PiNJu< zAL*IXA2po*#@;w>eHoY24;(%Qb^L6R%Eg1L2r6$FCk?$-{jgvmNmuLUOzM3?Db2q@Mn8wVau@ar+Z_e^<``p&dy+QI=WG*;8p@T z!q#Ci7wg17qyggH1leIvpFTVZvsmkZe2<@5EbQ@1u<+mX%))OqEc`!vW1;nBTox`J zJP-^Kcd`f1BS2>&f8XBRb5lrFyh(1sVCdlaA$B`rGfK=g+mSQt#Twmsh3L*K!#KcI zo0&C~gQq``2AR~sIcG0NWxcLNZ(PpxbtA|@zb+2v+R=~(It}%l8rmD&>wV32`L`&n_NumwRKrvNDXIS-x>z@ z^*J~?)QimUj`<}c*>}s2WA2!(FXM9b{u4(IKXV{Bfs~L?3HqQSWU#H`mLFe6hLR3w zqU(jO@9zy~`vw8-KGZXJUpCx*pf~PXUmEUqogn2uCr6UwS~4Cw^JVKUNGys>(dpip zJlzxLkn*aXb?Qh>v90cW5ZG4o%pBasIcU7&aqDyu5pd{@>UyY2)akVz&GHg4b)dX} ze{#hNB2~R&Rtt!O!#A$1=tP3pH=^v{cRo9h= zC+By%`1Tz^-uj>n==P+%MwB7tEo@6~9Uux;WX!gcVxvyHeA+1D>Lwy-an{dk9b*AGEN7tx<7vKC2dA+;*brMck{CbTVAbAwRq42R}^8xbh zR%$XYv)pK@xpNpN^QUnt%>n)4b7+=WGP82AK^QZnmya+D|EVTHCJ@Quq+U7v!>4;a z9tHkL2(~*|VV|x%3Vi8+9wvnwS#9CsF6y0fGaFsm>7d2}tj29)p^uv#t?zekVbIAm z;tO;JcpC9}p?)(N@h^whQw+j1%Y%H2q?!oh4YfdI0NwrO3mphUKm!zCL|nq+YH{(r zsFQ?y*;JYlX$Bf0`1$k2Eu=|R3h#=?)71*U4rvKbNLxlaYK6~gF;lb96|}Be;gy0G zw|^z!gpo*9E8IRyG25GzfZ#$6f!j}EiWjsd?(kMKs{Dt9<3RcMHOVpVyz#2^>5pnU zlS%(K35kU)BON@cv7>dQfX}*eoTpiVMAImAuvTqMhhECl(0R@spOZqIUqxzp%n8x< z%rIGJk3$5>6P2R3-VX zv^e2B8}0=1c(EZ!!>-g%$M>39<=dxOn%_+S^kieb`gu)jGHE`vN2*rJ6|z=>;|`Hl zyvO0q3?)Mm1}+P-64g?p3SU(=$G4)izD)&|W)4fzFgOz)#8{Fvq*>aBB)#@`n!9&P zIAJ7Gm85sUGADNvIE@Wvsj)#7{=}UYTcDzvY1WmG5Y_|bLz)S3wd5Cv-l=TFWODTj zdpP+G$qrp-Q$7b|Hb>MT>XS7nk$JYhIDKhYJVQl>6>2$rAUipUMC0$tPu-u}W)p!d zeOLuZ_SLMdK-mVAa9eG2$faK|;m)m$E}$NTij-o^olz@Jbna?bBCoBjqplJw z?Dayaj`|e{p0%kaB~L0{+9lW*TW{y()jef#ph!Wz4jo^o2!$+P?xSE_nqV+hnrnr1 zD<2irE?1jt3hkvSOD!$<@@gPo-J=BQ)5sNuqYi3K7x1{bUU?`)h@Xh^Dm;m0uUrUd z+5#*QN6P;y;7$`%wI=OwhHhiJoY8{Iy5XB4wc%+QOE-L2i+dlsVeq@OOII_)T=RwE-gc%#2@B*kDIs927WFbt?H)vCjY z*cYMtKnO@6ES2(_sX$5~fL#ENwg{8Lg<5F|S>7tm(io;TS3APtt9GMI^h&+uP3iqx zsKS4yWkaY!H@^BhVy#lW;)j;WmV7x-{;?1YcTa1Js1-*&^{=#CDIEC7YoN`O8qIUJsQ2Wtybz65?l3p`Jt{e9fs73S0Wqpvo@>%Ktu1Jm2h@8djIy z*O3%li>akZ;J*zG;?+R;PtqjfdSW95$uR}b*PKhbOE9TLuEwZ-l_ ziHxW)a>Ti%x4X3vNenL3NLyV8Lu3%M>%e>#rVd)E2jQOhcrJ2i;c^sS#3x#chbv&Cm zriY}DUBbeBdUTFTIAJ7G-6f2{57(AD2llm=Q8M&7Ee;)csp07HE6vz)nP?g)uWA;> zmD>6_&6i|r*(}yF=P-{aV2qS%*u*oVgx9;BKjgM#2Ow6n_e{JyI1o%tyqk9Za4@3` zZ5@oPt<6(B5<372Rk|O6z{mV7?RAiDovT0Nn6ClHU9q~jm2Xc~^nR792!GHea(}9# z_aj0BZ|B(yNi{pCP>|Z{etN%3^P|!bHC7AjMe8BHMu|V`ne4djp=l^ZV{V7P#mmEA zJ&9O5{Ugkvuwsa%owoGU>q9%e_+dQ@FGx6HBvRE*M`q>8sq(_LrTEwWIGi6Kc9>^& z$JmPfeSx#~Eqwxj#``Cl?LPN_d()b3VImU5eW-^&^c=8%AOzbrglws>+y*cGs2)m% zn=j@ogjwWys8#W*$3 zHKa3015uT*r-Lw998&AoqL8+wi7B%hpW5(J5)?_T3zs|Xar8=o1vHb9+1%#krC?}q z16MxN>EJT;H8`KAgDfRMe7@eyw&S1GD?$ix4%@Q&6h2n;b#9r_Ip5$b6eJ+>9;mNz zMeHyQIvd>JSgil)_Av5JA0emIrS0K7gU4wYDAeiShEGk2dfwqiQ@Y9E*(%^+RrRwP z%Ml%tDNfTO7{O>Xq0D-?xYe|4o6BO}GHjSTi#Zwq$SDP_zG*C|k1FlYb%Mgpo*YW%d&Vy0wQ-c||=QES6YvF<`-k2L7UT21*`h zxSY#MRvlU`;N)$AO2A=nNR|LDWpYg`5qu$?kyymN>Yg2zcBS>!>xh9X8UM9k$@twQ zlJV&;YsvV+Ws?l!Sa03#+`GWpBhDIe?ZMPRyF)hMv|rZ6h_?F{=0-!D`xXzLDlTm- zs(2rBzkI2_zitYlE{>;EyJIK95Bm81mo~II=+A~WfrN{77gYSGU&kXQ$rp+>|&9K?kFy6 zZdAFf7!ae63zf0WS{;SV6t5WJQX?avBQ+i_AmdIae0S(%W081hrzTK`CUC!^C(whm zbvOZ`B;?_ruA=+#)6^q$E@vOg`HzO45Lqgqp)?Oni%; zbl@eS=twGB$Ra5vfJtZ??MWx<6|>dbcODFl0n+JS(`VgPmQ)>_cVqev*N+{?(oqle zvLq`@q97zPwP~jYOjmIs8u^->1xTzQw76xO#l_WX>G_Rf;{Y!x6t{>%n@Da!p_f<< z5XD%)vbBX%iy-^dp_$xh@ZMs*TC4>4K~Z>1nd3F8Lb)ZKYQYg?%UN0?b0vHH_>r7h zP99GSYfFyOgNcDU6JuHNv^kUl6D^Wb0oOA%){=5p)PS0~5ZTm&N3{6=W_R(QdoK*l zr@9F$q!Q*{0|#)(GlcX2qAn^P(C!VEToV>0>@#Jn1u-N z-&km@7Z!`>va8MJdgH+Oc(Iz_ES*PHTHK<~*Xk?d^lN;Eo1;z>+U7WvkxvcL2P|7Y zL|m`I>=JiTIE9SU9e;Cn#ik)+!8` z`(e)`<{=%)uVMcR3lixHX~wDx8SKS26ZB9DOtl3$4R8>ke-vX{dfL1@*6>>R#{9sC z(O#{Z-xv}X-mcNb-(BK`)t`C-fZrN}L{FO=l~7tx%`*?&&U0_|P-`TKj9nZz0(B%EwTTL9gJOeD2`t(R5~%g;T>wDYtR{hmRm-7n=BukmfdK8b(A4T!Z;3KD^+m=2fh_5 zxV{EfVj)CUU~(eRCYnA$(JN&6Z?3|>Lo6-O#TRtK`RB~y@@UeLu!K4}EXdn$#n8E) zHe8rHp+H!qqynyIYAkE-Xgb1ak(3t1Bvc)yi9t?6oKexBP0DNe(LsCAebUlLj%XMz3Fbu#8?+e#e+J-c{wyHCJ5~4ng8-B@J3uu)k2%o=C%& zpfw?YSU9i^!3YZd%+6b0Sw^v_ND;*=y6yyZ4~HJ*oFb zb7M(NCyG`PO9-6cnHP$6c#B4|jWs%I zSSwZePoO|6Hm{zDCdmggfQ02ZW3g$mvay6U)x$3$i*Uk?x>EH-;i6_TkGIQB48FPH zMhiWRgT*E^5(amZ4i1MVhL~?dlS5clRz=b+Y#74Xac+9l^?AD&(A;3MUL(wI3&D;R zVfK4wo+#Yd^xxOdWU!P|f8S%OD=btjo9|S>(}q{zg!9lsKNWC2Q)4aYcldRJg<_`` zo?cxkyVX@I>L@b&*eLAq8}x8rnzD9_qt`4FLoXsJOE!O=NRg(_^WGZcw$m;n*H){8 z$@RL$(lnrX#-aNxwWxVC=@Mhnk{nhNN)D&7)>+h~11||hk0xDsEozd(NmFLyOLMv1@QdE#SzPe$@qQ7WL)hZ5@P!8f z*P&?>MVk^h?l>fg_6`|7$_X0e8{>Tm?F!_ECb$6qAhsN|fkDO2jl~MZ4p!-g^9s&( z$fbu^8x$E3^>G43|132q=!cG_sH!BVjigIwq7xk55cI`TeG$I5EFd@5*?}mxQ{+=5 zsH=!U#HLXnVWF5LdngMTs#+j&Wls|8u!*599LV9#8SjNeS91G>%@0!WQtrxSgf`Jh zu#8m8QXMdypCOtHUPA7*)TFY0lDJ$x!{Ei%w)BFyECHw@ihz%*;QR#T6u(7OEtTrh zn@WTgkeUQ%XCQ{(5f89pQVEMucBRr&!$$h$^4vHbubah7Bb6=j8?){YL!#;G>Wmq6 zCEWHSBZKx+K>hO&)L8qdKQ;R(g%$6k-qa>;79pOO9;$pAPxWE#l4dW57+c6*9{4@e zYuS>|UZy!_5q;UZNJ95hY3l%s;Phc9q4dtq%E`jHTdO3!&s{G4XW@SN`ba|gyU8UN zt)PtxjroBGr|4zgQmx)){y*tu{!|S`0-MH>SF4|g53+>V3zWUX8YRx-T=$1?WFoT` z)@)@lB09N=``swDQlSG!s(6EZ^?X#uR)boo*jiZGDB#|9Qzq!5>d;tj;UbhvJ?c}7 zBS4r5kE=sVVy%rVB?c7U$ThC!z_qiV2jFH2rV0fYJ9jIXIG14 zFD#*Mm9}hP=KvyyWrJ%KT(hWcAPqjP_Uss4<4YC7bf^$1G#JR>r;Z@majX2rRzO<} zE8yXou0B4*?DNh*nJLZeJlh!mM!pC#uiYxu`Mu3=3!3~X(@_6J_i1(f6Gp2 z^E!Fj979gvM5Oj zU=o@hO{z65a+1SJLdk;()VD=MA}~qte}f6<&0F|5k3I0Ctr1JPTh_=MlqUr)6DrTW zdYgLCnm&yhFML}^MI_Pbu!=4k4Une4ij`J-ENNQT=?WP~da4yGAr!sDm4A>{G$g9S zKXI;#{he9M{waUb$Cb>vugg#x;6t^wH5f&r;!Q(UzNzWIJmqxkVmayqI1wl_YFmjN z^}cZ)_C>v#N6Z!emB}LDzQOYz^G<}C<5FRzJq3}t)y8*25iX3IVeVHZ4vk|1L=7O@ z%Q_mk_UNb#cOhpLB_w#3-9xaJ8gxsQ%rlYT4+a+mg$=M}G7>XSoqF<#BZp?^Pd)kM z?EK+lr;t`+@DxWQGzK$Ak3IJEsU!2p4qH#hk8;xUqxAJ)=J3qyjC_9>RvkYbA4J)r zWdt|il3pXfUTCi7ix*)SH?rZ*WFA1YT00LnQ2YB%fHYzj1X!|w(a4&2oxYYxd%jg~ zGMS~~GGq?*=H`h2%8NK(T9RL&v^cKTY2wO=={CKN5~r!HfI+MlAn;dr5VG@88Xg0H z25=rAT@?8GFy7FMb~J-~gTefIb!8C9P>4x$dKkZu2vOi9bjQ(+D%Yh!zQipMb`!;y zOk$uX^7ro_L0XdH#r5o178$1Yk7OShxi~hJ8yT6(aWP7IglAjwSx$URX)fH4={qEG zr0>S}J&>Q^k0%M-79Pvb_VQ;|m(8$lOG^_y7|w<3=`naof7Rru@0;hkv@|^5>C&nK1%3;QfQ(F zJf`;NO?!Glt#6t;b71uOd(Ts4mh8n5d@*)DhYz<#M$S_=l5ulj&K}(DJx}Tgl>t{* z>PXRyeI++$Pom27s6s9@6P*jxvIRB9R`Cyy)`gFsnzt`A?L!XK9nEcvKn~ z0Ov0)MV_4-ePLARXq$$7(zIKwETb(Jll<{eR-uvm`w3Tf65eZW2oL$ml@v1IK!L4`auPS-M2FQj$AHr62-M8sWZvj z;v}iN>Bz-meG&EcbXfrFi5A83pU~(Hl>Z3-gayHuK<>4wmY?6}Pgc7efJ`V94g{xh z7mh}3W1YhnD67X)TS$mOrDZOZP|b`TX6nqN_YXBet=_){`F|6W{68?{e@T#Uvidwm zuAXJMuW_eMvL0m#;illoNd#i9(GPJt1!n-lCU9}>S$X4ntF}Rb5p?@aG2#Dj@5*B& zy^8zZNAKaau?={Ae8yvYW_G6MST8=X<6ZCJ^|7;NcMW6j+D%V)&&=1`(>?CPGn36C zh(cn-mnejB0!fen5s0FMB!t8X2||hl|3pFx2?`;Kuq+M{F3AZBB!uKw^{)E9{`#8! zhS(C8)jQw&-h1DxdiAR6)vH&PSbr8b#T^guWN)K?MEtcd|AX(mZXS_l^ z$F3>`9G%mhVhRhB_MQOZZCpWohF9w2WuHffko8!trxAeAJFh;0fG#;7XFc!=a*>f&fNvU1Vv@yU;LKHc^0CsWETtnhRoH;ybrQzSDXMBC^XvCbCQ4gk}T3 zeA6vx12TImZ4_tb#e1<_V{qccWUByZqW9NoKsyMR;@7}8z-UCvt8L4+OeMJKAqpyr zd%F5kwGk|c;B*m(e^&f;tY_z}c#=N7zj&hSNirc#zU9z*)DFM(+@hHQT(oq6x5aev zt4$~)nWT8LV7l+aERvX^5GWit74bvh+?$szfW7Z~nwx(8y$)dt;iKQlaxY`}RqH=-ICyQjMv(_OnR5ex>*l_~|$J~QCNhzkY)ZNMx7i@!%FLovLc&c48`^A>h3L5?0cFp@e04^ud8;J6D)D@zj` zDWW&hNpZn~y(4i}n}cej#tY<@a4mup2Qgk|3Mt4gg%{`Ecr1y@5$rda{y_e3i;#cZ zb%dBYqdn1y77 zop}1t!{2s{%1j~g|LE;y^RAs2sj2h)@BGyU{x1~a>v$q7SOA?+hcP&BPv3%ij+wYv z{qPe3QZoJZ1yK^rO+u7NfBirtIjvhUvNWdmw(^!^mWg!}Cv1byKBQy zXS_Fvd)Bxq2X{C`=+c7V;1CN~5+t*QLJ6m=*eYXpLV+WFgg;@)rn?||bJzKdg2SQE zT({)qpQ7$`BNmIZ6^dd*?e(BcV+5toLe?(&iEE4v zGLy>tlg&h<;u}n|&BW1pA&U>Vgtw!8Z;)_r$beu-GVM&?qyn;cmAiF)9>rXUG54+%kOmA z@_U#`=sBXLn`XB;3cd_|9_;TNHbj-`EM{SNIKoaH=cFwN@ziRAoi2@<50l<9?J~U7 z`6JNvarLR}J&15xsUs%-7E{O`yi3*thBD%vzWq!A zw6UU0O^8{HyfyG5ta4!*asD##{L9~*z;!8hJLIX-@9a`LXb)A@=@c}_+Qf8;2`jEYzLU2B~S0q@%Th`~{u z&@m%6coZh1>GdKAGbK+Z{sF`RP&6k{xQ<}#FQ&UD7+Au_$r9GfISQ6_jCPcGX z{xNJe?^uj5IdFu#3K$AmSPVmfH30??x8c7gwvE7diwdB}V-aLU)~x%gArM@sGy@SI zkfrGLRHWCg*6fennl+>1HM`w_>sW*>ih-c>6o?PSPQd|Ua7wE^h=ieq#S%idTa}jC zEbF%ihlj-+#5NKrx+cB?Aef6_D?A5XC`5~q4!T13PaL|<02ti}m_w#7F>+JVaj09M zyc{%yq8{9w>6iP9zm()?_X+AdMsK$18ggeRXj*I;n zZHC;|Y?|y8NVrr%SQKSW9v+^l7N8HLy$A5G3InRHf>R>#>;f*&kP5(+RfrVm09mBY zpv)ZE5;9Qf^DNq7&0<9Q)SEF*lKI|#hAAezzkWa_AB;?KoRo{Hv%_FYjK!73^c!9& zTL`_D{P;nTpL)5!_{Wj4b+uSFRO~6W88CwQ5rdERcorMJ*R`Z2hq2am zV!qsYjR_RGHP-X~=xUlqvq(I(5wa{N{-uSe_!XM^{lzc$LIJ$d*y5)DOS*BF8x7I& zi(R(V8x7IY;Y(~ZK1z!n7e^Y)iYw_O3zO!gppKdm%N{AQXE<}6K6?q^E-*L7b#Q((+5M3T*XPOl|FWMv|tUnijilY^a>!|W*2p7n^t!z0BHny1BhsH|~D(TO^-UgR z`NAS)xIm^eoDEYSbiWtjW|j&mtLS<4r3zGI=9OTBxIRQ2vSC=O3hl1@Bj*7aZVJyL znh$Obi_VGpDyl(*X~gp>tyt~tgo7G;<~74D*vB!~R2^q9G!XgF2{TvBZ%<=Y;QBa{rr(~7EXWp ztn;ADS8}y95*2bW8#aeV^h$FMSUXHRvuSVz4surp5NnZz&{}-Lwsf8GvT7cTwIc0v z7^qk#O3Q1+kxB8_H?!VqDs;cE6}^{i@2$ryyoW?fzsr`|dq}i&bLyD;^v&=WP|Ol; ztWYVz$jflmY*cfFlvFy2UlaFM+>*KSnOJuswS`7pwz{3 zb!PHV;^dWx$EJgc;S2@h!k_hXoD>gwGS%l3&#aHs8Wb&q9M1h&I$UA@FC4{`;l;#( z%t%&X5tS+{^qBS@(^a%9YAr?Xr>Z^&PbGk2x~Z+e!*cHOvjR-QUOvj{q+OUXirN< z486yOF=Gtxrnd58G@?O2Ik{k7?=OBnvdG#qSnV%$m1F#70rH+I&Y)>3bkF!j?~V!g zELiV5f4q#7=EmH7opz^{MSX5a$xg|07%We{4V@gSnqzuPyg=1?)_eAn@?;UZ3w#nd?(l4NUxS&S-??s|qs$gunS#J_;0@2>iR zV~3>!x7J-*OwXM+6W6&_8@m*IoRYfDTZ1%1Xk1719s$m3zVfsq`NKhrk%2SpG>6?i z-#KIJ0E|PRY@EZww;h+S#By-7{hF9hiW!J~4cHGwXgw^K#D+vAf^Y=_mj*7{zW-dL zzpez-O%9!AJS?CdP7qISyh-*NgTcgaO1&Ki zl`FW8%hwpId75Kw@Kkv=Lf`ZTF3=UxOH-E-NNr+hfHHF6a;bIV)wiMbW}EEZ)+tSa z%Q&Nv5akT444C`+gFG_CZ&P%o;=YONhvUXKrCa0+{@8n45KT40>_Yi-f-N+)hQ!r_ zrB=R0e<{+uJ+IVBbZ&D5Xg<%pa=WG>0h2B2IMiK&@XUxqW#NJUT9N}l$J3lJp$jam zPyd}zTKu1#;GH^q(ux}nPyC0I#NUnWSG4pAnaO~B|^!R~8!^;TcJD#4PK)Xr{;(yAzqFHN6bY=0zJZ;~J z_DW&PFgJV9kZ|H|6NY7b+b}mTKW`ntQo5h0X4-3*8#704JTcx1gYA#Le)!g^4}I3^ z_T+Qu4LY@sR8SKe2f8R(W2or7i)6!ho<-;hm}(Xm=T~YtAxZ0i5`JauR5mvcotG~a zK+zn-iPqSo>MHUSu3^{4GBXXolpmwm6JyX$7qKs%P?0ExSi7cf`EZ55Pv`Xu1vn<@ zRB-yJhwbyMkNawyM1GNEg&jpMJPN~M)nd2Be&gCOAQ;J7d%zl#M#;ybdKI^(~S*MA1*jcYH#7tyo zW-HJz)`B6(sS!u3qUDHf*?yTn^EkG7Shxs7&d$ce!*a(C)74;&EUt4|6(_(gdHpPC zu|{bpkpq&paYQ3z?cyz5h=;iKbw%XUV&OEs0wOaJVhx+FcFSR#1raWE(;Y5H-}8~a zTQ3B{wmxUs;>gvEjmh=QI^fKWeLZ@Q!ER?V72sd6#MT~hI&g|;*)5qq|2P{}*@n@qt(MYZvhbB7qL#t4LY02)-|D8ZcHt-oX69A+?QjfQf-V zXl*A$s!yECD*KliShCUJYF6`Q27gp`js-#32w=}YC&iOYj48GqDP&-g=6PmZJv7h8v(EoHFY`!S#~s&YyR6KpG2ma5ss z)pz4B&K$8iMY=HY!ao8IU?;umYgTkS1UXx2@CgzMv_uEaJUuyp(?QDjXRM;>6ub!s zH24~Wmoq-wE~$<7v5BtG0UrSm(N~r;7b)YRXcbxO)dq42W=jpaI7116sbe(^H6++c zBMY1Ez3edK7Q)GufqwrYLcd{~;ATK;rjh>+BmT3I#_hSMwg|pqm$!ixd-rh`JNu$o z6+>=*7c=xYb3qh5)Y**55?%9SOmAUCT-h|c1}!mf!btj;B$0HJ$w%LUrUYOet9Ar> z3lXS+Q$2)1`(Bcu?UpM+z!5I*IOxq_MP|(WcX-GfA2eHCwJ+(5WGBpQbiGyuTx)3D z8Tj_=_39*n-%G#?mSgc8MBl^=yRFn>F)7%Dh%tU>>2tEKhc zFpRNG|Fn1J9TtBMJV$%e_u;_$32_~ah|cI>$oPY6eSSGtYS2kAcK33xmG<;~4+{Il zRyQB<1+lNUSIg`kUx#HBr+oGsxY0H5=jQFcjlQ3qK}>4C0Aogi@QkTrehsk+O|t21 znRaW+k_T%Ac5sbS9o5cNaV@7*!&RILR69IShq+2&7CzVHdreRvUo?f`RSw@RPPxg` zMvYf|mLl&GN_101XRJBX#YbNaXcQda)iGW$pGt&p#Ob6lb$;uWN-%bf2d~&6#$dfO zC!P6tTUHay-aJ#Ql-Lu54-QOlaf(K|tTt>|SXP)RG3cCU3N@r!rodEHf0?vBjK@&L9X97V&D$T@vV$OBgOTT zrB>F4J`h1%S2Dcd2*V(iWq90qwW-j(02jSG3AdJHzf)T_Y|z=olFRK2S+hV%7df-< zLVPCPX#Pf(A~o5pt5{9mnMtVUbm~QE@pvpqz=Xo|r!89i{&vxL>M7djXUT6!sc%Ib6G z)=yZYYwM*{{eo5a|5W!AwUqVmrQ~cC!Pc7#*g#nQR6SWl*w_1u|6rl&!XV9-(O>+5 z1;r_mZp#zgWiC~Ws{0M2WZM{pQ~xN5Y0mp)?>FFDxQmo4E@s?q`nZS0B8hlwAYmgfi!dg^JK|+@iNONFqgYgMdI+a4| z>r2?Qm*m^}X;0w*AzZ8uwvb)%$_1FY;2luV*{(^C8N+1_dN%{D4Oa;f9q-vf3f&@T zZWXukwHIdEJF608yAG=q;|2Rtx(MT!;f_{QU@YL6GgWK&lpTuhL?|CQFh+!msm~(_ z0PMg^h~z?JFW?xnvf{S1k0yx&H6}NInelE;Jn$ifJ-?U-F@bFp1n%T1y8(Lg! z!m7rG2$z-2w$E6k5mgEML1e?h3@VPmI~ismh^E9$6bwXJo{0m2_b{KrW(L+{{X#|A zhA|n8V3sV>K5_@9kP5B=&yN*#JWt7o$g+gt)xmMH)jTniLKOovgQCu(qfdbyadIAt zAHs~JK#Xy=kfnfB2(8UgNWfpfP-^u?UNJjyqU6_7w2L*nH6C5IjNr52oRu;>dkHfu zu%4G13uPJsGfUnwkLj;_yP926QH5!YFxcfQhBE| ze)&{(*&>-wAi~(BkXjl<+?B)O94wT%I;IXnQ2bT5PR!WY7p zlccphA-*O+jJ)UwS!7A5yLRD12UWNnea|HbQ|omv=6GYu!OG@zN7tg(1O4-L_n9Dx z-eTsV{VEmE9PU@d2w3{>s~<>?B_ZnsG|BYW4@8Ii6}$qP{`=|&l51i$05QB46=J>_ z=A99y(IbcXSb`iT*kKeH0hTN442jfwj11TvDipjjRF1wsog^d(@lQ@40uECW14Y-I z1_7dkKHO1*6JPJOWy~l;C#l6x+zhQx-~1kTA7Y;WPd#B$JNUNJ+S=Sxb+Sckp7(Ea zuIa@u9Cjm|h>Y-CiH=aI$BwkNYJIs`^Xt6H6KEhyQPU>xuVF1xJ1Jx2;$0N&O0ezV z?GdSsDUKWA)yS&En&^Rf)Lq_3 zXh+H`j!RQ23P-A%&HLDfqRs@N4!bPc!vP%LorzEB{TFOlGjM`JSF_0LN_ zax^GQ(Q7KwYgZeo_q&ah85M7&uH;RyBsT#_%*z<4;#i*fhS(;ZTMsf8gI%Bgdxlpi zlW$$8H70;ww@>r~_hU)(00D6FefIbL{V)=LO7h+EhQwW!DCGrC3wl_Tw) ztSph3=KE!y94tvwlO%u`4%d>5(;^a;eoDrEB{Oxa>fugQ8$_F1C^z_*6G zNQF9vLla%3b9w>Q0kkwir7>8&fSuDw4#S9DPYsrQIP2hRq(yVQX}&L8gK-z)nDm7_ z!cRkef;SXSv6WqveXRj2(gUX|jXJVNLfW%Mc>xLyd0CU{0QWas^`jHIO2eFbDfNl) z1mig}y5;7}{xgk&@ThL71w0v%*O4uoM2FD>c83Zb8HdY3t+`0wUFiZJa=L&S8|wlS zq#pJ}k%CdIp)tm9zAyBICbLsaoug*+)W=E?V^5^Jo}m$%!tq2_rGO}*4*P-WO?~hP z0ZCuo^fUhBat(TGKZggWd8ro;VmO#x2-ef0y$oBCki0*Jo4>1X`AqzG{gK2#jj ze`jq(F!xw(8~q=;18TaV8+7= z!7efF0=!UhM;Nw~Z#T%rfMA~;1hxpB{d1B4cmzN@3KdXwt{VrcG*bp!h~5pH>cIlZ z*3|Fz1e~^}wwj&Wx?5NR#BN)}02*iHF|;C-x5NFeVlmqO`-*`hWcd#EhIxjWx@d8D z5_divw(7NCVjU(i#axxylmK0p_6EetsQCMJw+wj`akij)~Q3wQU@7@DzV?2qe(VygN&0(Yuq*Yq1%upr7cNTP*MRVSim9RadZC$lt4r@!~ zoXRJj>U0(j=@0Ckd8t4vcc+*~6TrI@kwGHefgAm{C11sj;uO)CIx78KzZ4+gNvOk~ zAu@{=a6g@^0uu)@9RjEyEoe}-AypGT%TEUzWABM)J%SpV&NEr&8>zIh>dv9n?;ropKILcv)*4I>ViCwMC6gtCV!GZ9~R$~WM~f5M-` zNsc#}o|vGD=TA}^7r3&Mt;*%D)nT{6TgQ%qzdoj-agUvP^msK}TWHkAW=fTru?4g( zR5=p&Ib403sZ=i3#*lNG&WUPtW{_UY&x%AW+tlbBJqD%($oN8pEaNy(Or61QAr)Wk zzD?Y?Tspzs5_v|04;MN*440$h7b5_-UWW*(ebxlkPH<*yJnhfB#ceRmi#skBhpAw! z5uIsP^d3vIXHT2!8=zGa=L}v+1IwKBX!PL7L1aBrk`;rk;Hkeh*A104Sy2I|@*f!Y(EMEj*6@*-x?o#{|u($V2swoiY8NFO@N2x6Y+v1_8p zB|sA{J}qkn7SDGe=EW%rb-`vQvXGh(B{~*va?(S3Ee~liM!e-TcWQTAufBm> z3g88F2(%X67fF4krIBLN5~ZPgX;q&w-A{B7Np8-9ud8TRl;BKnitJ{} zY|(I@LV>O{PmD&RNv9CxsxeboPN~EcNGs=8u{YxYTPm_1sIlfX7Rezre7Z9YKI(GI zQOnu-d~@y#kFPfCtRGC@v=D&PLbr`d;f6kqjadqv`)pI-3XWw;u*%@y2N*qVQLkX}mp2LU?+8SBD#Vu-;#aL1}xU64Srk ztYM33P?}lR&5O3PXcp_viDt2_`9s6d9xGJI;0UIIE5OLjK*2oWM+WAm(lEgP417qM zgp^ynNz%a1Yl&X1XyZo+i#_mzqSA;>?s1 z_d^lFj6;7+ovCDTAV@Cho?~DPuiCa|icdz+(-i~%reoj+bv8&e#7!_u@(3J2IY^$irYZ%410~cr@-LJXj`-$GzN~tn zSnJ9I6NhkL-m`S=PGn9$dX^L&j^%b&S^v*wq#AG%t)-OMT{@fe8A1v!IlF6 zsR$AX0&xOBD3msR2`)R_JqPR~j<{I&5yZ!%F%Q=87o3LcDi+Pq{iG8j(CSGFD$Zkq zLk^swIl2v|0twT~bIQ{1Oz;u-=nNDPt4PeOOrqgH`97LHVt(k}N&%c9lXBOA1^#1X zjmJ@ZOq~b%r#Z2TljserH#*J;tD87vG?Y#F(@F&1*L)*_hOSuMw;ii9pJrBfI^m_{ z#u_Bvb!W7-Gi3iLNwl46)&Tlo7XoPscibP!c0alX6I3fo`;4^uzv>vGn4(uXATcD- zo{I1bzk#b=TMe#6a$Tg#E!EPT76U;GSA%VDp`>WF)%UGxzrib} zf}L}PGUB9R|BUuKp^-!3M9_Xb>{6e}pzn)#OzFo)wS=cV^-3;75y0pzKDQO)s;;aC zSIt%`b=)~Dh?kFS4E7)cZyndwGt0F1Uq#8@8Up?=W}x}=Q6swCuS(3}V}{i2Oscl;+I~(VZ(F5UJmM*DU_iehGg1SF z=7F0Mjwe=wJtgc48riu5P@Tu~rb2l&cnz-i!}yZP7Zxd+Wf>_{2*ulwC3zNLf)M_gxPmJMo~? zJ-S+EmM&hdKD8R`T`VkS029Sd$T7UTsbmI$8Zt!GfoBeAmzI{&i!1Q{DW@w{q^T}E zJBFIIvDM%jYC%0xi%gy3cw)t@20Ktbdh7TdK+E3hY%cN{Dsi8=-h8HBsbG2fh59Ui zhO=IS+z>z%w4X;hxkq*cH-MlPGImkf;b-{t30WGi<9Qa z*+;CneYKTxJ-eKth!iEd#JL)5DOHHVcVnlCh4f5=l5nHtYxr3`W=a*h^2!*z0b5Lj zH^ls^^@JHC#*;zoiw#N{xfe{Q?$?1Hwa(uSBB|v3@+_v>Ec~m872y0JTyG!UGRb2t zuHakqg+d*@U6U(SvKKRjg{)u7AfFHMR36{vLiyR%sm4sOL$LrXJx~#1-gDTB3_{1w z;c$fJ{7QTXR8a>f1a5FFQiH#Aa8;7PW*hBetvH|em+7!ze|rFpq~eC#LqGMxu1USrJv{S#?QOx=ZAOV=Lhuj z4tl&yKhNBQpECVixEDW9(9fp(@$Q@1--^E&b-z!T zWYm3}Fv+Mp|9bqzsC$?&$*6nXLHxz2`)|S|qwYO~Nk(0rFv+OP5GEOQTM38I#OtB?+o?0(|na1T@~sdBClE}8GB8%Fa2gikODK$RT{*qSA$)M1X@z@ULeW0 zAJygW!%}(I>KS63S%j6zV3pJ$N7M%`PY)IEvc=BfDi&&LIEgjb$88l zHM+Z6{lLtSMy$w>NKVUEa&UIs(WvpbMCq4bf)ZJ5T&{a_IX|GyjmKaKx? z7}K~NaCBS#RtvnTJ~Q<BFd_%M*EN8avnyXaB5O{_$-lyQ1!ySpvNt2ZlHp>3yA zX$9T2O3>}#Py7bM`(B$ku_pMr?Kn9+4;oIlQO0WMxQ#OS3avc{ELidIAHU?*9Jk#> z59*pL`>V`79mlP$1uef*skfbO+bcVMgFCMJUaibuDZ3rlaT{)Jlc2%a7?jqShwqEU zRbVEYsqE@4j0yRcCffL2LF5ZIk7DiK3=wLthFhN~ww#40b7~Rj-}Gdckns+aZ&&YB#$2JPocYL6?7V^UNiFub>!l_ayIyYR{^y%-xZ#;#-Mj3y z^3Hk%Le*&oP?E)(=d9!Z{8~+|#+N->967(<3|d%bC6AV2w@s_fz3lomm-@7W)y_J! zVZnJ3KRRxU7H+qVfjSk(bt+gw?fIaQqovddn)Bsgz2UgcW-GwD!jt~><($P*cKx~o zF5sciNDNG4=17%Xb{idsmuk=w{{r-idl|T}+``J{4`PiI3iRIqYh4I62D9V0g_sB} z_*^LfnoUTJxRq~9kD^CPjVM<3YWOY&513k(gdt(nrnl_^)^xn@w?Yy@@Rtx3AJ0A5gc8i9`yU}t${CO=P}#MwE$*{Oi;YJME7 z+)GfY&iOMX|B~OCXJbL=b*S~Wc*sJo5VY2y+-Z6jW{DCxL%Pdu%MUQUQK@Lth9~us z=dAid{DKZ*l8C5~9A%#s3#bB?sRy7QTC_ZO6&4C}1W^{B#4}7%V=Y&5VMRIZE*Qpb zi=kFqz`yNoI^ISTDuS&#XQt@Yf^|*aS`b`v+>X@*YN63B5t2^P7ka~k?XpHH1|z!- zwB$p^fO^8z45O0YDAl@{IpP5D zlKM(+ZqA;dxjCb6=EqvMM*Q{0>=ahO%`5PW%$W3 zeP;@dX?kr%z3GRETMfUm6c)vk#eBPpnux7ASSVrECkYz zH5m?_>tU!_cwe9IPe!YQliQaT#%V^G4DKh=7`GPW>BgA|_R4lG?>Hh5wqbW7&N-4=MFRfpG>r^cJ`1<4Jg zt^r)74%{|+iFa_DI_m*^uNIy;32hxIy?7>RoCnE zjG}w~s$a^%O!rH|VebS1JY+a>>mCrmzwJt}PK^N-q8#gRFA>2|F062nom|Z?w&2ym z1td4D-(`cD3S8p@2F5vC1_B_^STwX|b)xQIu_F-{i}BUs+8oPd_AK}_8gOikEd9Td znwKnmG8ilkE>ET8Y73LpL#4svP`brsWNXTSVsIEOG#8b<+AlPp!LoY>@hOB*Wz4kh z`VEAGRasu%f)8?s^GiSk>dZ@9nuteh1X*tM3B>h-v zFgcO|D_h89GF&QrGFVnNK3NQx2A{VNU*blPk7yVJ!)QUhKO(42kKXNyZ#aL%RU7(! z&*cU`ZJ_1sz3K$J<@df=!6pMXtD+M9DHU`X(n1ra`ZFY|rmFuO|BTwtr``0YCC=5) zatE`ae(sK&Zn)u$GFaPrA%)u|pRy-dPWPF(&?wk#liPe9?AyrQaO}0Uu~!279TqqV z`W6gvF%Z-aTkOd1DIxa-yKhiZ&+ZT5^k&8PKd?cb@k6p+-u`@Zs-~?&Fkz+9qkK8uE)svDE|W2~CJ6 zEo^_WPJ-@|dvX|<6?6NqIV; zGGj+D0RFO1-voTn=tT9rbJiRhs%cd4K>Bkxy4@+MkPQP9(T(TG~4K~LzS-KWN84Z{QCm0T?1&M`GCc~w{=Xf?k z%|a=w5z}CGIK7OuFv?=IG}tUE`IcPVYqO#^LFX7)jYXv_D3ubaQjQMW_DEll@u!e= zbfSY4oi?1<628lrWMjGsvEA`f7`#AAssQJza9Sa-|rCc>s12d{0y&Vc#T zCNdQ_1td+R6SECONn*8N6B(EsyQP2urSBj!rV(^7$vKKhP~6lj!#Hy5e$7XawL>vp zB!=|?Mx5c+>Oq_O0!3J=WIPf8e%tQ=DUQvGU>dzG07F?oPSB)uwXQIT z@;T(}pgHo2I55qLMPQ}CZ;;VTsiLsmYGi9k!!|FiOBj!^|NI8!z2(V#q>naR{$*f9 zulQ%T2sT0_q|^yoL2Ok}O&CoPPM=eKfls-s*0;Ns^L+#<|fWriYG&uir53?S`^Y}8Ab_L^6Q zX&l2a#&Eo=fB9M>GVU$J$h~ky>!* z!YHEw)8KSG8;NNlmDPx8usWKB0Jm_;X1p{Q9nVIaSU6=hVj8Rt$H=NNm79f87Ne!X z=A9|@tc6CZVbZ|Al3LeWSY$9*8eEQMVRcyOWHVkGjE-eu0$C_!G+-K>o=$6!S?Hu6 zD-9+`v#<;;Y_b_I4MxYaF;gv^vKlcBR{624X5)X4*DnUb(XRJ%LtXDrBB-i%COA6J zjwKuwxV}Yp70K`yRs#Ga?FNwz4u7UpL*5U&g7EO*jKEbDZ8;JR|ABlwK1@T-F`NYN za)2W*WfKr52B$sdoOHs;y$ROHKZ9d~b&MlJ2^k97POCsVuRXxH4vS{J8a;9K&} z0nvgG#eOd$^fqJ#Z0A_sCY;ECa+c(7zrcx{W}L|h9JqIsIR=ld6~IB}Gacf@FGhfz=4ZtbC`E*QgMroO9rEjjf?HRWx}`? zLdUWhFcnUjEF&AIYz9n&)5%O1wS`;81E<09U4YoX($DLc;!R3Bq0AQt@j{$%l(A+C8oEvq2VD`grAiP4vWUlF`Jv9tAgaYRzOFT>1ZW-%8!wP zRNeYw${J#$WRGUSW|m~gW;`sZR}(1Om^R0=;dl$9tVSFgCrjZ1hAb9f7^_T1OoP?I z$>o8CNt#j8Kz~fFvs5CS z6CnulX_aNoK$OaxLrC{F@|2QBz?lhk9S!Q`-JXpoLII1E#_$6Qyk9l+A!?aC&466TyOftZtLT*I&+rtrK;!Mjgc$ z4~=;oevZ$7LqtqwQ_-ZRY+@y0xv0sFd5cc?kp=cSV`?ef?MDW4`zb#B@p15;vWuEV zY&c!TWi-y!kJ3kN)ntk5XkA82f$3Ie(r=Y!qrzGLHVErS_~MU`gYBDT-5T2x9f9RG zHkqv1l@CJ5$`X_YiEV7yDiq`|gMfNHRsGyJ;NPhxOWVn#fM+w>23|af1HQlDPTV~0 z>OGOhD7QdKH%=N1aHb>!&%i<@qXE<4^kh0e#X=_SNNKQm=K%XWiMwQ>kZPDT@E^&* zWwY?eWVkf=98T{(TIggkS{iItQhTx%G8qh(3YXMAwT(;q!P4MzGz;(ELMNN?(qJ?< zMySD91>^LJfo^m>SKD!PfV}U())DMSo~7M%S*aNzQ0(~jZ?PiyLg5r#uz;;>+85>{ zCJ60f4<2!_PLQL5PgER*DnWU)eNN?&of8dbuH8j=1{qoi#IpU4Fx4D|dufZ918>+t zrkYK()l3OnwD%2@1R|Dp7GW&QAsmIx0X}0u zmpYXFwK~V=5S`;lA@)9+e7dN(nFDnYnhq!}N~>X1@Q zn|lW`wgu&oMsZ)>+1sIRLa?jQy*Y+il{*pp&WMrN*ZA~v5u^wE!&I8?1ht+$`Ay2* zRpelI_pO80tVM^xZQ{rf<9(NZ$Cw!MFgx#x*L8}Y2fa27J*WlX`($<;ogYW{TJ#;S-^4V+G4QZ_v;|%HEH?812Jni2j$C4~rWd-AMl|E})oM3$irC+;3Ei* zbZ8r&-7FUkQS}in3Q{oW3u`!pBWp^M&xec`&g%i+qa?1p1&{J6RB=iO?2VZ#28yyxFMBY=U%6Yb{MUa(yx=n z1R%gFQ`LVTkw-(FBfeelqiQ98MRpxg!ms*!)Nb~q?)N*oU-kF#r?Tk9qBpC$_tWN5 z%Y)8w8J=*R)oz0e)L^lOOxdE>S*N@?98=)+sm~vO*4f*ImwbBAvuh5TUnFk)=!dOHx~1c5BF>Q)}>R83Us)zLuw9s(RzLaKvLY ze}7Ez_g6Ka4(+7)TlY)K-=8{CL!uiOLpUrY4WsO zs1X!p_QjDO3cpeOjv{a7x<5I9xWbLJ3&kV3f+t@$2PUU_n6Ri~y=l$JUkjibj{RDI z7NYmwu~~CVvF0w>HKFL${S0dkwb-ivE+Vomqs?5exFF-^DwPZ2>=ISS5ZT8fU74fd zvcS9MU1UU~&Z*90MY5?X+kheHY@A?t%esNR7SEt8zgK-!-@?%=T!D;BdT~4pMqGnR zUQzW@uJ%ofi;8&9gpRV|P`GUwB4-rqi`EcDc9B#oBt24cX-%Q>Rlrepk8f*KcA*&1 z{-B%FiUaFPT;_lvHK@b|#28{>Q$i!rO|PUq28Pd_8Q@;SeTD=CYM-y73i=Ago|SzG zVy;?INm!`{V=7dOV|g+HPFF1ACQIIASSNFKosY)hOrI>pDPrF)u8&Z6Fru6pa;k|W zdQj*+y9LP(#<5H{P$QzxV**wruoK2OVaI(LVJ7CJ*c1FNWsby}2Ty@DNy_u#rLZW0G7jb`gaM5+H&;R>%(*$EHHAvW z+W=+FMX7XoDxTFI&W7Ni2!bvsAqoAq76dq)0R_q$hVDy;#-Vfc6u%+O1?CVQ8|~85 z)w(E}DbKL8P=H08TM>KgIK_lJouSbXoRJ|GfZ?2c-abEp@nEJ?Sjiv(EeGBDGVhgJ zZE<#ms!)=8_Fy~U?p(Iy)vZi?bc73Ku7V6WpU>9I+?-O3MqNIbfRL7GO4J-;XDy=W zsxz1~<&XDnp@~r&5aNekPwM4VwHX;>xUVEH{`y|UISK;WU5(sX)(2gj8dm))bNV*) zV_F`jVI+&184Vai@8K9(s(5(ssCyi|tC~4(eWKWM7M{#0{?NZQi_pDu)vd(b z(c4b0s0)U>Tv@MdP_1Pz1cF754_^QEYEk#K08NNeVkv|d^A06m)P$J_dd7O@8OHn3 z#8XFU6T=pEZ>N!H=3~~w80bf9A1ya%4oM!A>0AS|uu z?$lFg6uBiu-9b`Ga7)SqXCU^GBLpo0sgL_E#NGo=0oG(Xm70VEEj7Waze*D_Rs9wG zGs5q$(5)#yqGas9;tnRTdhTP8@-NEjUUe5W#JH+So6eNdCDqV>^873cRYQ+*;TY1S zs-hU9h2T@JzKOGsd^FOWRMoejL9bE)7QT&(lFM?>c5(aw;Q?$CdPUq=sZ4j@aRh>|w& zV}V!>DR^DC=JUyB{vO403k)cIj+4`f26Q;iC-X$%EUT>@dE}9N?gfPH@_+$lzVps; zRN!urJKiagCyqhMonH00NH*2-rc2U*9l*x8P+LEsgT{q|9l7JNXzfkcXxBjV~yPfxomw=+Wb;>VJz& zOCk&G$CPwX6BcKIsdtuj)pw@Jw6VyEgCQBLRtwXN28blmM6Dkm?Ef$fWzBPG;Z= znzD8XDMLxYMiA|mXIZn73Ry+{2f7rZgX(VTMhW3O%CSfvn{A#Z0>n7wPTN6rrHC4q zE1S`9vJ>k?Q4xgtOXyT*=d$lY**5(qTu`^o5{F{7Co;pxVv-C8#Uw6m5AX8DE$Wyy z$1-)0etS5wuKk%>#ByFRUSmzq+} z&Ayp<*7^doa6fB|Uf+6nYklwRy~#su6h5_3j!#z_p3HCtw9v~CBQ-av{4o9tmlv1I zbYM0+_=&w-zTxDO-w_^*Q(t4x1vzlkrwNaOta99g6Ra!RUX`&8C9lfxV4h%{x#xi6 zXVk>kWY-aXklLEp{rrUe)~Ma=N!`!l1ik;+djzJxlH7R;Ibxg6(*l$ueJo9v=jM!Y ze@|>Di*E-s0KL8~La(tG+jm9=NVM4gsiL|1JTJEGLK#_bt0yA?y#*^_S+TD8VR%ix zoT~mvcCg-P<$qP5fbSav;%KR$=@82c-$&E>?lb4efyN;xTMOeD7lOdND^5z+*aGLs zNXF)JKvhG@J^=3&akvQV;-IQDwr)iTd=*2}_O)BD(FLHgv4}OpS9il5q#p zzT6Nv)*& z8A_SpUhfBKi9e6D#0?>fWJHWoj%mNigF^wKfr4Mwr2>4BfF&>3PgqCFGV*7dvfkc% zirq%18|s+Pk*!H_GYXj*&jI2O!i-{{$T(?Y47F;d!(sUrm#!hS)}p$mp9NFhbL(jT zB&N;nHhnGd!_Q;J4~xA7cKDtMfQe+@&nm`OptH<-2!?k84$up)m#{+&nxmzP2GOjQ zv8BrXcDIbAr)FoKE`C?METxOX30l%{rzO29EK&AQ`4i`6aHXd_ZSQs} z9OFV^0v+SRaUQ9vpoMS?C~3B?x-Lhg0Lbyk#Kw^U?=nEgIWuc;TDXF`s`!95nU5Q+ zU;V%*CdS*6-bc-dlhcs`{~wNUdz^U})8<$vQ(z&K(ST!PWEre0sTsn;CxgM#;PQ@9 ztf|*_Z)DRLIE;836yW|PTLg#zclEdXPU;Ld#rK?FX}FXz?V2jb<)G9qp!6+2E?@zp z+_79|rh?oK)lUgG^T6P3o&mvs9$5y1ZI{ZC-hHZ+Tt6_nrrZ}4{y!~8X1m5Nb*Rez zsRn9;Y&324*q_m34~CZ#E9B`Z-R~z8_LC|6x*v)2=w`*a2M`U>8SJp*snjyk+PWyY zlCxa|lF}rzr3NB-{z7bOfGEez<(cQ($ne83lzWdX0^H%#;be#ij@} zPm69wYv(8sV|I*zG3>%_&xUp)W*6?V7ug5k5cx+{Q==`w;o$|xYQWS0p znwv(cuCK2bApOK!R;nWPxw9ZzE=WJrXxKZVgDJB1zc3W;e}_3+PE2fQ-dUx2M?$O> z@xM{?vLHV+%7Mva(?)m7C+nm%tnN3Pu%Fbhx}TX~%W3-iNH27y-VCc$@;Ufg48H4t zwU^R|M*W52ZPX^mZQ2;sYh3nm6DR39Y%yGRakPsP>1bH0RLcW);Xcvk0Xs`Ax_2~( z8{u$cg=co8ZnGe?R&eMs=K4L6xsKCbF>Q{7wAW-fTTo>(+}PMydh4l_^tW(HJyaSz zc4$5@Yur4G8F{p%`?qSf#x8`PKi+M@gltjWjl6UIT|t12?LBShNC3Vy;;ckJ3v>yh zOIN1gbO)NTSkpP43kWZRp}hk3%Nxa6I@E#PE@Js;5}&OYx;{c>~F`0KWzDU=j9W01;HZ#ddGy294vg~53=i}C()m-}`ss){%IQt=wksCS z72JtxZ6yk(L1H6J6%G~1tCZ%;KCBt?)iI53t3@}pl5|Fj6s}XF>_)eStH(v^iMo*$ zf{c>t(TRc-^d%%t;dWN)rLUCIwOlxepsLTJi<)8jmXz*BK5t{(6lcp{n({sMdaC-H ziHQ#pyi18-vV#q&w0!-FnwQ^@qq2PcM?$`SjZo4M>1zvah$TgJwUCUU9TYCr6ZUI* z!iKIs<#6bJU()@CFSc(yf0jd5lmcihCXo_YXm0XRmY}t^Fbjc*3(8=(z!jv*XSfiM z)`i@vaT_IAV7PH`bG{SIli=f63~igp!zr}r_?b3#J3&zEUm!UHk6I=6Bf|yuQ_y(Z z&ZNT1j@b4is@Kf_qt?uga%5gJFJ3J=7{spHCqN~IOL9)sWcaG)s6?Eo`z7VX?>>Ky z<+I?zg9{5A{L^Maud#rnHh839N6W+15mzp#N2kw_EPyj#wEM>?lQ930J%WuqzFjX) zC2$OwICvFz1X8^h=mjt=j{PbF-H7n*0#3!9>Zh4N^#t^uABQU+Bq=@uam}88pxATP zX3y$1q>DlLT7DHoRTw8{Voj`vHL(&gvF?|YiNA*~0gZ;zI4N7&E;A6H4cygQ0yn<; zf$u#78S0$&y;nc*eIb^f7o{ArBQ3WMV298~%DmV}!G4O&$|QToSS2uLIQ&DxnE2Og ztl@r1E(fN9Glq}1F{p}-NUbN<1lf^6fu}=WRDm%RzlU0oSJNWd{@}<$UA_CCbs^p=p|m3 zS^+YfkwD*}_OgKDOvUhj&LE@~f&IvEh5ZzhbY-07@D^p6M<}J&!yl>ju&E~y>*13j zIsO_Wq@m1f8ca-;qG|H=q^IMuo{mK8N%u>*p3dd4O@4l*+o7ZIWwW4A$G6fga&YpQ z6Q|ECxU$78=fu)4Fp{hzqtEK`iS!1ZF zl*a_+BDx`Vy6U-nnjAIY8l_s7?{BFEK@+OQU&Vir6oV^Q$TcPfgLr0}7OPM<2x-zs z;3`x3Y!m^&P5~}6Cb+m1@!+4l0PV%OLy<-;P4mF8pLvG4{-JS}rK6VRAbRWd`JdGK z{C>?4Sf9UrO{RmxwbyAv43&TDn&5gJS#PSU<~F&U&}{KPX|_ltCv?Ara-#P)&!5@s zi-z&RJ=VBziBJc_+bC2z^;*b;)Q&2di0Dwn)#D}GR*bU<=~HUOZQ~DH;dJ=`YwcIY zS!?XpwBgrKaB2YLdb$0HT5kVR(-q6D5K@QAmKcnfVcv0j0y*{NZrVqyM7#=s)Wz0Uw>A4B@QlVUW8j!-GZ& zhvXco(f=cjej<+4{gQIz=6P|m@^PFUHvC8{J#qo^VA-_lVy6W`YuF0N?z8d~O4uLi zz)vO?q#SFPaj@ZlW<3o9MGeM%b(0*j5$$kpzzCAZa4Tb=@U4>)2!z7N(_R7lRc znt)Gg0w!Wz-7hKYuC73w@oq3*7t`*RDs;N-d~>sdYbh&+rR4(=SjjsM?bLwR>;_Gr z-y)n+#eU_2Xn^22jy6e_+|}C^|2@mD8`J`H^dszPg->pk@F`iVf1Byok1$dzGSXn; zVt?jt<(}Cz(7OGubyC-zcUW_+DvKUTAjEV z8iC3!Jb?xd@jC{Y(Hn`HR;%N&weDxwdg#Ygd4IhDNtqQQ1Hkm_y+tHa%wt+nlw@8Q z@;HfQC6~x1@}#wNVnmZjiit|%NRnc*`-CWmKo;!_5Z@@d=afWc6bNxHyQla#NmDa5 zNm^s#ej3zIl9=+BinTw(9ZX`%x&IM2BfJ+!qq)c**A%1zh?GT!{1OCQ_>uygX21yt z)FPy#)2MJ*a9-3k8T|&Xzonho(*o~umygF$nPXi+kSr}|L{&^!$`ui5^)@_Yd$w^E z4R&5p?U+;|&JyQT8cVpO6o1}nBZatzZ8TJ#64`VB5a7}cH&hprW`i5%olBnA#LpZ~ z>ru`PY(B@^#N|gFh4sh#$ z4TLxzD5<0l7o$d`E&?ey6d8S@uB57)L5HV-gDxS-(KV`{p`*|u*AMMlxnc{|rd82K zy@+%F{S!Xmb+cz{sYb_X?aW`*X2 z=p~#x(UR_r?o@Ep1<6XxJjk9!{YXkEXmc!+^=AQ> z(ST!PWMwbkZc76TwD9vZU)U8dVytfsY7z6*aX9R+v*WPoB(l)LC|UcI_}z6|QqOnf z3fiefxP^rEFb2}$YsTWR>epK@ohV_#`6wB`|`Ysm`+FknKhDzmfk@v(hV#yu${cjE`7#^MyW**1 z0Y0dOHIii@_ZiVqJgorK-EkA3y!iN+hKb_e1E|j?CJy3MkEuA-c(bj;KQTld&A+;n z;a}6KcfVv_IMeg3`O>_Ia?FG0s&u)46=~URS~@L+L&7l;)jAS%Q2~l1L6_-)L6^LP z+?1lL9AP5ztPwdv?M>teb5sF& zFFIl%%{jHAm5)vl0}j_%H92_@xbYN55LNB1?TAhV+DKgy(HB|B5fEAjpFxfepH^bW z-m5pUy-C3z8jgZ1)}bQ+p&|iPVz1TqYO8s^T7pt`_~zE2+2PoOhj;-(FcKjkPVmy5 zkst!!OMyQFZh|9|kt;1wZaj3sM^0g)6YzHRj%GnjOT=U|9$4iMB5OZRC&#ooqVz?I zd9~2VWVo@hvHYzg8St2mPZqzV^s(2d5wP~tn9+^O8DwYOWee1( zopr8GH~pdnVi}uZ2z$~==@ycE^Hd~cxRSoRR?Gh z3r31#9UEjgjz<^KO6kStKZpUVk(+{Yh#}KSZ4pT&DUG-;#c`33iTiybl}c6gR)oTJ z?o1AOD0IXd3IT$=j*HwObwg2U`gD#CG(y`+kYs%zV=5x|q3GJ@TxO9tSwqq!sy;)& z!!se=t&>(c-ls;Kze82*7)8D<5WCzcRtwoL(k+QVLcbug{UU@qb_?1$8Wyo-O0~9g zJavYcDYutb5C^NcdkCoG!esybT@Uvr5vmx3PgPoJ*L>?ytB=+qD-@~;VQ$<{#2XeZ z%Ix5(ajw(ANld(IQDr`gfX~gHK6)JUzl!UUXp`M_p2wVQh~F@uajzqv`?qR96-YU$ z+SD*+Xio(9E@DoR0FJI!eb{?q+0I(H0~%lvy#4lYz$|QaZx@WKOi=aHEGhK_BRw1$ zDUl8SR<)uO9BhL>m|&e)*d=NngKL~JRB%`p`~hVEjlCi*7*h-~HsqG#JDP<|u+Ye6 zyfhe1ClOE<#L3!E4v{&1ad(_dR7)rRrCCVv?*Z0(6E7=xZ}%$itxo`J2Q7ZZy<>!Y zwg@`1>EyKB)q7_OJz&9@YM3h%qfiQ@#!yNTB*vAMVu0@)0r-ms6ljD zoZrIHEH=_9m8Q1lp>zYzfV4Ya-I+PEa%R@yq-q43bb%@>0ws_z*&utzMcZ0R zem=6^t)QK z+qyK5IDnE1aMn@9IY3y|_~+Ic!Ai4+UkkFytd5^5QHNF5!TuahD#PO}`Mv|+6G2t- zW7*eLnRU;cJ^8$ogDZ>mOeZUwA}3G#PI)OHsVCky%d0#H)+;V92s!*HeB6ONI2*YN zn5tN<;f`VoW|VcRorK#iK>XNbbh4KCq$6P7|3hSv#;MZ2HpYjzRpO(}2mM=Xg`R$d zul;%CYePlC1IjrVIQv>6>{Aj!fyxr$P%7rI=$mq&vB_*%EWtJAPz&p8{^=Mxg$-51 zN<#VQ-|^hG>kZ*p(O_<~jMjKoio0!2@tO*~j24-CY$GyFr`|ivZlJwy$gc4`ohLB$ z|E5jT$i^XVy3qT2efN^64Ynj(hfgCqfV1oJt~n8PJyCCB|B0>$IZ4Ia5YvM-b{Xd$ z5cuPB?X=@1gRA-)lk@}W*{wc7Gdfj0h<}C|nV3$#_2i%u3;Vf)wW0WBs3!<)ssV`q zS49R6C_3K?P!|bhE}S#6W}!?4rr>%` zUAP5{7j?T_lnE$$h@`K#+-6gSVNp>A#-a*zYeGU#;vheg6`(d(TLE9nqN3v*lV=S# zH-`z62l!q<&UMpkQ}#KKMEurA5gNve?f?)0X=uan=A^DQtR|32LokU_0*%;@o7)Lx zH|P14(-#qO#0Y~ zGc+Lc+Vdw)(cctIr_u)PQp?B21p3q|xwF+lbDC6RIM=S_BwMoc+I|;H;+cr7i__yV zZCo`QSB)<%qFjTI#ZV%wLhkHaCCYS(d7}kSFXkma^*19#Po!X;P(n#f2iq#etG`sO znYb2Tig`LK+xVe!h&o0t|&?I zVNKBC5=P!Cr2BoV?l-&y>0bz)eNd@z_#7}pYM05JL-3^TQKjBI_MY_**2coSrL%4W z458y>t%$KvTxQ)Xwh_H_p0GkaI1&yUi*hudq=R$_hc!_%%%>|45GW`I>y29A;`kZ> z64xK#vNu_PPo!;0gNPEgM0M1Ha~O7pb7VoK0v(qrQxqx0b~$*5tUhJ!7*ip*`b6Mr zLW&HQgsNZtPh0e-OXhws!i-jY@1EG{QOjz;TfJ5|2>j!b(Fb#FZ+i=My<7fP^&g`T z+3ib+0UuF}ufSv#i93L=4~J;vde;qvm?;}AkTn^JvsMLvXNI#UkGKBQVJ2iPU5DGP9MV=r5Biosm z46=AS-`7*&BFF4+P;j0>7!!^M-HS;lWLX@tZ>TES3Wtm8oY#!tVVPzG97)Onf>Dd8 zbCd@AI-OI~*+|~kDVh!El1lkPMCfO$Qhp;c`d~q3RLYkvm2ym}6a^-$l(?{-YLGmv zr^o)4@#L)TXXNZqp{^=CX#36>+NNLcX#@b9oAWAgHxE%XkNXux;7-afeketr3f#@B zfQ%Jq6k~;hfO;$?W=Ymbp{mAG#ygdjh#X%~F}Wl$$a{qLkio{|vM#1JMdIITZC$^yg(Y*qqx6nUhUF&he zhD~P!)v}_u?yyi&vb7 z`Hv;_$Fllk-xcS=1=&>$5c^+@vZO{?Qll)XQI-;pg6;5&uPm!^men}RYMkYy<6xM5 zYM6a$n0;!PefzG2F}i?^S51)%?}sy4)}%Wzm~@Mv+Ft$tNytR4O|PUqOs{czaIYbF zs-%zCIpuXu0ap&?6fL~iv)eO5miD$c0@~Bq`Y`^&Mid0(DV5;!5J|n+lZU@oW^a&~ zm7>ff_&h`#7Cxnpi~2UM(lW5z%N`2uAl6XS1yU2X);LrT&^M>5`{k#Vjvs7IkV{JU zD!uUYeNa~)#wa>7>0vE4R<6A7m3c7Mg;F`_IJO*iPN3CLgk zgFZwms|t&AA3)L-wy3w-2G(k32d>rW%{@EB>sjc*>FUh3dK~w!9Df10R0FJ2-X{Yz|@RPoOt$2`8|^W#fQiVHZK(-DNR;1L*)fhEv$=A02E9R@aIDH%`F zdsy?(Tw+p_e7=7g`N?=#UMRb3YhG&zro+I-hH4$hoNy3K|8H3H5SHNBYlEfXScPm1 z-wp#C=c+n%x0dz4Va;PXR-hW?Tlp0H8qdA*GfEsuF|@!!_(`QNUKh+l^$ZkQuM1|) zxx@JOb-^5`*@Z3r^diL258qz3WIH0_Z zF|%No4`P6>ehY@u^SWISLjkLPhC%&AW}upQ@qla&8^aCx*&;c6HC;2?%rV zK^7#+yR=|tu61#HFx8UnV;AR1n_KRV>tj(q{yiq*GZ89=C&I7^Ej@6hc@UnApNR~U zh;P2x@Qngv6yKtl0G?0ay{A4jMdo(97tBe>?zBf|`*jg&UuYPd#$F^X$cyMbH-s#sh@ zzrtK|+nd5Nc5*;RbB%@HACAX0R?nndlNWO18L!>;8VFFX4U{#HEE?S;rAfG#Az3f} zJ@{rOLdOaDW-&5MqUrdM;Tr|UD85nc;`!!QmNt(pD$dcr!@P4eGhj#aj)mhx^0Nu` zl+`ma?+ED-=7N4bh1C?|ow!~$@5H_b@3;{q`+LhuQ1KrkW=F6`q~{ zE7CC$Lw&npDD`1xsGAPjm(0+wk)S2E&i*({K!^B(3ZZz;ms(ZRQ~wTA^`<)}u+U1QH6 z6EXcq4b!VH4`zDRBA$y9GyTCV0Ug5h!@&iU^BgK6qJM{({$ysrj%IpmGTtUXo6xkh zdM0K1gNIqf=N-;*cDj*(hx564SGHcN^i-c@ZuZxsMbsH3zBry$YxDR<3v-PFl1h zp~=Fu_noAQn(+yWQeX^gqItgh_rz>`2mTqsdvxOY*)1~%J?>&A<-zPar$-20sgfVSym55m%(vM3h!vfvxxIH3Ec=wLY@PB*6Nv* z@#q3voaS3^(>;iqzgCcXM%&-0kVj|r#LE}ncWT=3-0gNpYqrAAF<}=X6iy*RyvR_A zc+TKDmzi`mO8|MGWDF2NR=Zg#ao!jwly}= zK>N3oko@}%zxP zQshs;ac|BNsklWRg>^KmSS0&@HOYn(Te6DP)3D0W530bDWpo>F(e&)yM^)YC$(kvt zGqtkruN_|ER_U$3f z&5{77S`kzcNZi7`?&2cvK$Ih->V5<>)w+``h!9`7Ga2SL`bLihC0e@Uu+Z$jR1_X)!T8ztJO~7Z!0|%@*CSW)phjhy@B4J zXy87b_(q>W(OR9Sd6(oB=0Un?t)aK+;$CSAvcmWvJ~wPZP-d`QAbp3dO1w<0APS^U zRliYUDzb;4>l?6_!&N@Di#{QQNcV0eF&aTry+XZ_zxudN?R{B}7X_?$3yl*Vp|_LA z?rb(aly9dBObpou4KY*l_t@+$)9G>r{)ggUt_G z&b)TXvZ4x<>R_L!qEhy7?hps?`7o%{f^e|U%@%5G&uoahRAJPN*Gc2c zi{&!_D9V#IH`8LNeiV(We~((41@e>lDYC;=MZqwgsvk##K?~zk->M1r6LK67jOD}| zNlv`G+ZiY*xYgaJ)!}ZVT=POgrMKfEN*!T@_SQ>Vy?Yu(aVh|G?0Cf>xI`jrt9Sb; z`P1~B>D?weUKEROJGj_$>vV|%&dqN1@Vkr#_)XPdw|cLfclH+cELY2eW+b@@mpRl*4jczaBtP^)&SJ;z}l>D|VHPK*(jfi%#+qHa@F#UV!Eg>_c-d(ke$X(d>kbv#Y5Dkceprl+gZb9b`6}IXHj^nUld~NZ)R_Padx_pTLCjd z8snf8glq#xxlybb))2&K*(+iDf-bs{^$!h!+78E<)tk(Of%Iv9OB7>$*x?!LGalOvfhuNmuE+}U{(Nj-Gf_Mu}fr3)} z2m?BmC7|z2G1nG$sfHOzlI?=26j9XAFj)#&k|oaLR1>T?3tZZF%XLhfM<-hd8oAzt z1H@1@ELLne2bN|ciVuc# zmZk6%ZLWpFJLjU>Av+E53ba{vQ>FCAK*E;wXfK@sAP+9K78l~+bli~~hubBGS$-csRBkpCe35cL631)t)H z)IMlX-EL?{>)zWic+(9_E8}u-D9<)_W<9z#oTA3KyOs-z5yW{1(9(afye^=44AKZT zF22%0F&O|2y$V~VMknCgxV;N*(oBo)okG16qz2E@eeMX`9e=VNCGIrf;e-G>zE= zp1ONN6Zoy@IC13b)!jF#Y7h6wak4joA3VQUSbqGmr33pAaat@K*oRQk-o-}|Z`!;6 zz`lKpkM6sU*0Du_1xOo(iwv^lo5RUqA?`xVQ=97NUs zZi6xsR=bT7MZBG9Ff(`>_9UCF`nliERYV>0PM5;!mK_9H5^y$ijU=OOQAte7hip(st@+SbCk03=v*H z1I+M0IiZD*0)fr&yZZUm!ZPk{3=(k$qWX$uARftJ0a(CgGThkMsMrT^K9a#=wDHMe zxHS0epHMJZI7~c54BX*+#cswLSL=fR){CMiUQJa$Jl4Y41Nn}l1IZ-a&-K%8VI+uEJ)rFq8CvM3p8(8u1&j{&1GGSr}8sTj-75 zH3gk3+chEB?2C$|FoP*!D=Ddd>LF12mo+=h$j@a-%>$a##?^ZkT~iE_g2KI-#G2R8 zF?pXD$VPiD)`CvywQR(AExbC((4j*PD8LKbYb_GJ*FFqge(m^_8->mMH4oc=h$=R) zdp~=SD}ryF_g}(hKSr3LKl@#t>j!|IEnYZ(0Ct~f%i#ak>qJSvu(2{UQ{bleYm?>o z;c&x24^bdbd;oH-pt_ zZyUQbWs%p2?S+OXTpqb-$hnBR6Xy2YK2IWd7VX^*(v*0woJPhc)_uC}iDLBt$7h!b z8xAzI$SNSGW19BTXglpCIAn4g0*;T@YPLM=-MigRPhEP(XFL(oDln>c0i@= z%*gwJ31-m)pM$${$=h5<3KT7HY-`LydSh<@H?*t~ZSZm6EjARyz6y^Y@%7-Cz&*hh z7G1~(m`Xlfp=Uv%J(hQ6M0QJOWHuz^^aJ;($cXf7LN-bQ<83Z7KzId2y&7q}-Szji zh)@8qSMz4AQVmyGh{wJjiG;MJL^XeM1ScDs~;=F4k4Gk{W=z}q&A8mqK+?CPY=lP=37qGhGUYFd9MV;4UraDP^KtdWu)+B4V6BM^#=on1asq>K_s?i;@F6))_6Fl?I>2sG;8U6cVNE(T zz^=%4Eo{l&H^4jMB=t=ZoF&r@CzvTxJir{9kro`{?``L?eNK_`FR6pbcBKQt4)Z)^ zl{5nwa1Jkq%Sh3n4A=H7mBE0K=<8E9*B1!AYO0^7>6)tkg697Osg|6Vw)7Rv^eJV{ z(}`tG0iNuB`p@Hqy)kZU`hhfESepC>h%%8h`87?=38cy4H{7hY>%Jt%nM|5&^IEL{ zN!i;xkx72rnaxJW-8d$mZudBWc&oP`1y67(61F+`wKMaCJe)GmCr#$}&#yjK-tR6w zexSVX!19xbCE$35vpkC&?@r4tc}2H$iE`e5bn5)xcP>9xcbwxsoC^N6RNp;(wB53k z994E&#=rF(|5(I$4m=dBDFiUobAI?f_3NF>i**Y(dMx`Mq`%=O*$C#gTr1!Fnr7)) zQ~MwF)Q%x06+c`PG5MG5ReGgM?AnT~IohBrZgwCAk)l@j=va;rY_!KlFW`8B$FQA8 zr*Uz)6iZ4HkA1S$mW{MYSi->|$X9iIi2fv(dTtKeXLUM;V(rs`T|1l3480eAEoIBY z@pVg`E_>7%D=1byoZnur_y`X84dlI((55bG+mS}0YAGuF5biW@MBwIXbv|%c?^|9I zRIn`j*9bt+8TafO=6df?^QJrl^FT}4o_U7x?vIQYUKj&B@fY@KX{q30PyD{nG>4eQ zBIdAGaTwY+J5+O&<|AuFMvpEmjfIPJ4hP|KDr4R@^v`SJyhj;M*wrp0OS$3&*ZNzGUjm=?wJda*n!$B6=w{$!Jx@ZKHJ8o03<`^)nz zr|lE;;vMH#=y&EgYbA#*QgeJf$$|E}PSWkisPag0QDLwv(G&Md&qsBWNHJAiA2B6# zAcHC~wmztd*x@eLS=e8C#=T5FfKn+U9|YOn2+7j&08C$iJAo^pk*vJZuFR`;Py_wArcY zYXN0Fj2Sr`B)rX<7R6$rXPUQot$%j}%;9=X>Vl5pGKB)`wKs&=z(OGMVk8o`N)A#8 z=-*-B|6!KEzjY9|T6hd@6N6G**jbUw7|Ce0l72a8x@0j}3@%1)MvuIzqhJ-B*4zf=o2n1$9DnF^ zm+B9qCQlW5jIXfHZ%P7iB4joObSc5MBiW;{*j*NxDCW@OhqLlZeAEm(6W& z3Hx@#(UCf95IV~w_Qzy03)P9bk0irR&Bclm){`(X?$44KACav68d)r?0w(d{D~8s4UD!r5xMpS`Lb{ZhRDWudK?la{gc*;c&&qUyGn? zE@a_z2|XsvL`$JFDPrn3x+-@Eq7nxbSZ9(3 zCI?Ye?a(DrTU>`%WQtMCz?)G->au6 z=|BaoPuhlgzK-W{xYr2kkk1i+MDjH~qS2e>-u%4>bsn1fb0^>m49tdUL;fCS0P@ z7}ZR#=-#N70Ow;7a0Yu2qw8dTv8I=?!jk=@rSZAdn!HJR$Iv-^nqvoID7+XHQOyHG zKJyIY{YYX$S)St{ik#K!Hp>-gAip}}$53^&t#6Saon3Gx#@3fL5g(M1{E+-Ia%*v~ zvw>Kcl3%gnD<=bM%oH=uOe60!W*v3r!bHffGZcD?SzCV|JQ-D#iL~|qs+oBLZT;$Z z>Sg|Oa-1j-S!(OvI}k|n8%P3xLVlTbTUh<5B|Ng7CSLXT3GS)t?`kq6wQ~RDvl{nw z-2BQ_<7VV|C}zf=?IFB;U189Z!JPbOn29J8CF11&qNi{IPQK|?J)8eojuQnUQ%=T- zFp9Bt!OTcGgmI{g;Ubh8siHeWK~pl^5TZq(7ki4Zz(Q@DHLu~d+)jW^L&ROFEy4~R z)YtV(ZKsJMS~}Qp?LVxiw|dW8d9A&f{<*Pov)Tn5Y-p%`l)H9r%f_Spt<`DG0u!v&-Fr1R zJRrx30+HofJ@*iN#1%fI#Rtj6YHS7FwF8cx?lSf5B z#O_VOIfER50PtYzC{(CVl;)G!4LOmFuyF;o?I z%O(hue6<>cCh8u8{b(Q1BC=&QKa2~5b(xAk%P}KdIODI?-783PxKQ-g{KhLqA>*z% z7uxjZ;wu{)SDc5 zIdowa2R2?=yz)we*uGo8_=S_$FiNaI%sq|A;{N)%DnJGK>6l}4xFor~YRK@2-NE_A(0@_HdEa7TjI`TkPlV!5# zICYn<|K+!=8EKi&v(Z$0;;{QtcbiTrDK^Q!@$| z$KeK3)qfTlX}Fq}DLAb^)f}QAVKx2eL^E&UJK+Fv6bYs8+$xz%kwpL2G~vbbt78#E zCh!a*#_cvETR4WFj|BZ@ECB^J{~ZST)ht2YKjBidu$y>@7(&D)sw#d{feWzx(a^G6qhX$e(Re6JG*B*t&SPxY)z;0k zkzk8-^cSxvp!M%C$d6|R@@QwnnlNMkh3u_!$6-$hf8)hCR|BP{;bLs^e_zIbhq@HF zF@PJ>e_zJmBN{JpWBTvQ7ng=N<{JpmF#PX`@nIw`kxR5m_ZIM9FU!A1cU;sR`7iFc zcqNJZ@;XKPd`)?Agqn)v6b-MO=dp59Iz%75Z+m|n-LkG-tOj-d)m?l zurGmRwLtubtX^$nE>GK5wh$sR#3-HkY3H)5ZDuQ|)DkWv#7RtgDo2T`fmZ@9I!0y-0RT2U5QK$87Eo0!y=LK&kI zMD~SPuP6g8 zjPC7rTuG^^@WU|D-;gEJZymG-EYt_Li6KDP7NiUnFy)NiMlugHQa$qw<9)|i*r@{$ z1Ev&3^N<~Ds{O^TUJKJ??PIVWKJ6;9j8+dP6BQOrNqfXV6=pPH`fNl@A^g@e!A9sm z8fyYZtf{->M4kCyFaCWDDv1faQ_E57GuFeHvBNBz1ItS3t5&1^NE54PKgFoOI2PST zFm7HS>tIbYT4uRc@t&?u|%!5)U@*&O!A`<1IQ`)v_e2~~H7q6DiSj_QwNhEHlah-j zr6Z+YBCKMz-8bB47?{L=6FN?Of6b>V+V1;nDGH*FOZo(jNe+p9^&uGR&05f@`9(M~ zv~7}$Tc2{tn#P6ZCUv8ft!Aww?lWnZPR0GEDO_eRuv2fqt??0wk~ZuVELRo=S*X47c(IkA1yHr+n62W6nJTO1+96#}{_q=fIpmToJUt4H07=>zc zjcz6_cPfV-+q)N4pWV_ST$%ooGe7^N-YcE7yT*m3?}ZVF^tNN`8S}DLxh!|CE1swp z@a!)TZc|mlOH_@pqBM8hjC?e>T6985W|YZtm(u1y)FyXcpes%dwGS>R^qZS>#d`tg z_7_m&Qbr&+Gbw1|)KD9>1CUMz`Pjl`Rme>YaIA^U8oH5ZzUj3XVq~1HwuAP9@{u)XmsM-lwbn37uVML2Yj5;@W1{uE|9n=!Q11lr}65z<_ z9{8<-a~SNQfTmLpC}Tu5`#64A_BI4$m_bTC2BL7+b#5boApm4X1P-@Sxd!ZLTeRT; z+dLwBO&C_P2JIoBI4g3G5wml^QpififiXd9j7k*2c{xzIL0rW%DmTZGJ&6Jsc*;RG znX1S$tV9=|xKnw}q2WzX%^9^aEPrWqTP(p#atW%KQk>t`*JfD@qV0mAlqaQrYC+Lc zPw=UY$fpJe*>D2uzDlRlj9=JWBq5g|AwO>E5l)B0v8cFKy!n#m&RzD4q)f0^LYL*oK!1Xp~tVy{3u*A$$deNc8OL%c9 zeWC8Ilh%nfgUz~eF!%dMd%OhIw z!1JK;tOOTtCeo)oVLDJuDGm+P9;eC{>}Ij^<-?(5 z=Y_g6f-EO#=l)O6!j<_tBcLTQzb?_oVtmV#3W|%9Uz3YgCo5a$_uJm?m-0F(ZqM^{ z`MVzBu)Bs+OKl!Q{ki)XWMkd_4ln{> z;Ab`jN2T;U%<}YXVWmK$EFF04aa?Tp_yHV5UOKQ$mmD5=eD6yd@GDG(i;{MSoMyRz zfkb?3h6ps9ljP!GvZ3C!d!76;<8RNkCBI&}9BY0W5+%BUU>}znxYKt?iy~z)hK144 zn^(X1n>qx4P>vG?V%(iRB@1H9=00>2%I1prq+7(vC$cv9oU>T}Q1?d93sp{`OjtCq zFUY*W+w$C;h~nam5W>ss(-sj_u8SHA9Ex^vS0`CI7+ne$Gc21K!UUo+TQ1_wSoSax z)W$y_ktX8^!Ee-Z@Kl(WqpA#boBQy+p#cU6hm&Oh7~3$k8_eYd(WU`oM%Y{et#%VP zfWknLabu_EB20@IbK99C2?P}9!q(Lsn^$vK&c?o)E98y}613t5W?&nQQga$@Vq>GEhg=L!vIMJr`Z{Riqr#rQ$ghuH|#*p}0E2?&d)JYR%j5 z=|)q+8egRJOBt2c_oLz^PqhL7pS5b6HN16kbr)s;EifEVv5unOKb&1Gt7cS7D7g!I zt)w*oyvRu5HD@ne%K&)wmRnW&u7ZROfV<;Z-NG#X{YZRnYL-=Gfd16-rpGU)s+(Ej zusp${v_P7u`$!6m<1K$AOKON@=tNS+BEmSmVsMX(IVe6;p-8mZlTfT#fFh!AG?*aV8N(IXIrNO-O7hG?<)Bb z=H8)h<)1vF4c!OjI8h+7bSryj&r&&dca{7>cp;Md?A;(V+@Y)@xEJK?AQ4Fhv$-gZ zEc8Zicy8g#$+?`HW64j^;uD#TZVAQ=2gT|+re`b8vAI>^q;#47t$C%VA7QXlk-@?f zZ*!03+uU?UPq_kye4D6nRqvENI&_)y>b3?>x4IvUS(JrZP&%)duPOhYsI8_Rue2Jd z7G-jqRs&a>ugN{ZO1tZ2P4-neP85htS6bo4kdyB=k#xwrKFIZ>Xn~K&LrguMal2{K zR=RGO76H+}%($h*Dgv1Vvn~i{$UVVDo7RpEX2#>}axx zDMn(-XHhObrqRH_u@xJZ;|&T?929e8|#l1uLG;l~%<|!T!NJj`nMgi-LbLt^@?q&0^(y!-IZhOaEak(w4i*bTvO{$y9vhuA8h^=DPp+GKW zp(40S#XDK@bg3ykW+cXnvSMwH_|h&mE^5%%BIIaOSlx#pJ~_R(S5{qmR1{=FP`TOV z^X#ER>l8oHVdk=jqw&1+i~+gSlgI={DNn!dV}H7alUFh_+LvP+%7H~t8T<0>_jhps zO@wn*K@VN{ioR@JZ(-7NP5$g`V(MGQ$QYdMpK;t z96Ncy9Kyv;S$T#J>eSoWQ@fY`8S8d{;vr4PRI=fm2jde9M;*mFP$Iw2Rfe=0PmX=1f4(TC(TWpfAbA zt1p@1aaIFbzIr($Z?e?#iyl1`sZGsI?E0$&1swm)JN0Ib&jGhPn>DWhGkIoD+r#d8 ztATUCcv)zbP6MiX`~03&#I*{mE=B$^JBJn@wL3^2dY$0cjNp2ivAmnd z8jFgWMDlJ zmOUHu5?0I!O$_PBgsV~NZz+kn(>w`J#|x2xhI=p4nEl}2)znmgu=jGSG&U^o^`jV| zN1I;Ohf;~x-m^tv>nOSS*la@i`tCIa>($UydQHC z<<3M)q@gEsf+g~?PwPp&B*%#Yk?9h-Z$v}@3u9OlO|t5D5VTX(k7(qRhFL%L8yaK7 zUDk1hcib!rep@G@#nEU!_AMer!lZYxOCw)3_?!V1EVNE77o7Lv1TW$rkeE0Y4K-ZG;fPYjNAyUVojBaUQ7LsS8@GGP_!!PXSf?U5p%D3$@*-2ZnFjdl;}v`; zoM>3X=?ml?*4muz=K{diCY`4!Az2Xr;!4y~jV`3&=sw?56MH-yW7m;zS)_e55tzfb z{YD{o9Eb1gNH0Vc0(to;Oi=|gdYMaA^*yWSH~782Fjf3S2^R`tbL9iCv48*^@(vef zsSbf?fHU@V=D*%WJxRISCuWUv6n%~=;=RcP%{#;!@+>@UT2M?9JM8j~RA5X+6R@SI zL8mh;;6EIp3?Ej}Ax^4W-%?-iwoyNdeiZ32zxgg3<_2mLySoJjzMXg@eq&jFGO{cq zDG6ho0{p?qNW)c?47B~AS>p;4R#o$Ay@G*cJ&Per*-~g@c8sAwT$oG2X2M+2-N$H) z2p!eYxdcbEeZ0EQ`Wk#Jh87u2=zq#`LT}6)l2ktmBOnIm#PipCEvZt2SMRk2R~ibhUnl#(BIb1i zAr0M@C$MiUoG0oY2d`DIMp#42Y27Y=YE!?bo?w{!?@t=5L&2H{sbCEXQ4KT4bmf{s zCc{ky+rXh;G1!D;h7>)55hRD0J#29DAucb+B zk?vZ5Iny~fpXD43$y*;`zK*f*gDfGAOHq%4Iyy{ct-{zq!gJ#2V9?_GR0CvSbjzFi z(B4UMrW>B~yyv#KG8t+h2Q6fa)R7JX>#uA`PL*!&$%I9i<0D@;hJjZ+#UAkqB;NqA9ZgbEv5bST! zY>R4#?Hp&$tb0WVsckh?^#NhyHSCqtsS^)?%i=KJ%*yE#?OEe=uVC^tB`0DkON?bO z7$nn&BC9jbPL65wbXw-N5J^AQ*mzj}+M!gd%t9mOKruLkZId^I*0bJ>*e)JhfT$-J z@dpyjZcLl)LT1=w$|o^n4i&d5f9fYpSVwUahjpGvBb6-p(v1^C`{C=r;-8~>L8lOn z7ADz@mj_9YLE zjAYs#fi=r}1=@sSGf1Zy&&`$EbUW@0vK;F)IP6U-kChYGTFoN61LNoUj-t6aZ=>YZ zs7}V*T%}XTU$lLLzsswS;&1thg5$gx#*GVJkf9jb$a3UZe!0%Y*XPYup3cW1Ip3tRCya{)o!y+8oVdahcEbvKcRi^l_$$!N)R1 z_Kv&KE$GMZ8v|*WNp6N1QK3Qp+emQnYO4CcJCX(ktr*%M-ng?*Vp;gbyo^C0F{$4i zBMYss#lDXDUYKCpSe7c1$zO(O>h)Ch$t=khTY$mvSKPUaTS(F?IK>YGrhGY7eIZLI zZy^zEQCR#QNzz+L4;e`5hYcyeoT{#73F#e5*(jxFeX;-DNDRgiwJ&FgM|3|s$`Of| z$_|REkVdxXV4jV{f1>^3k7r1T-gaS?Sm3KCBZ1#OM!Lk0WeM|Q=v-0~%gI7-g3d9N zNX#G?5HY)IHk*ak)%{Ky3hW)*t~5ey{Jt@ej`oBsr5oo7^}h0M1cXKSm?Ntc3XWLM zu1Mj5z0Cebqg$fTo71e|(w7qa33-wd&P~p}?BYBl8TmXIYV8)0EKi2G*YH|v;OByK zG-&MU$Sg1%e?&t{16vh0jlf4hmXO0q>pnih?oZ^4h@~U{&TCzUU%~f^@EISfD!0LR z13OgcfpV%`0R>Ppq1@-Q#MffdY{$h^|DO@D61Oa4+I%wM6ukO@Ph9&+Fx>x%gkW9| zPU)xp=w%vArloducXe0wR9Cg@W2Q^v2IB+>*-Eg53^pGD9ANqJ1F~yAc3EIyy|93HvoR)X za1uV&3riN3O`>0t9|QvXKj(hGy7k_xsy7-z+98^*_wKv*o^#JV_ndRj2X9aMH$yds ztQ?N;=BpIpwh1>@2n*28^`)ACU=BdM z5_|(+_8FdGF#W@^umGXDDENSY!LrlzC z?V-8>4b(zgKST{S0yH%^4=rkNo}=&WP~VB@;!B1u)ZADqI!4yZp|Njfn@)co3`Wx{ z)eN#gBzoGS>M zm&j^>+rW`UO#1Tn4$?8<15Xj? zsA5wBYS5NI9}ulR3`W|h^gWWviZZff8!4>gW|&Y!k#_|1n7S2~{9qMzkRWzYC?MDX zFpQKAC?H$NC?wQS#v~}LLNejYUs{kVx#6LDtp&bOx{jwp2o}RgsY4L{>!I!KlN!Qp z+#E#LwiQpoUwUiuZR7WaI*;ZZGD!QM8{VM?X5Mjc?@eHa?L@Wmfa`#=k+}mKS-8Ak zyKOpUSU#5~EQ$84nLsC5`#AVId%y+MV)5VZa{Y3eAl|IGjRlAG&@?d8eH6(S33O-43WEB?{E$m zwP>NI?$TXzzpF0|#hfPlJkBgU_WkL=CGO?dD=JTrHf@p?L})#`~K?xMXr z%|+|-j#1~0V=;StaM8g++AzJup)K;)Eghiy#*v{HvKDC z0fm;<<}$7Vq|gl^o!s_F>hww787LYu_N7JnHn^dh5r%d%=ViypXHn}Q$~xT6AnO7Q zeX#veju4?f$k`~}1A0X80Ukj>jmAR!&Jz?h zC_tMns7lBQTOo~tt8!rsczvRbAn61ni-;%MEWA21!ht?$g9x25;d3g}Ceb)jMFD7} zz!e%u4@Rp&#n0%<2=1s_%YZJT7l_(#k1!7;!{+$8E}(l+m9;7R`3n2tT!=@w_JjT1 z#Hv`d@0Hn_zPA|dmC$I>2d~VB@w-MVP_reiAncu^-bML9dfJ>wXX{xjkXDbL)$wsv z1Cu2t-xtXVtjWeS^vuDfklVvNMbSt13qu{1zLxBN zNOQ^dz!qY`#y>q1OpN9S5*)SGc!`>O=HqgWz4GW*%tHS*JVHN0XJ76MfziiC=Wf5W zV8Pg{g$Jy}XrAs{7_lZaz{j4shG-VDjza$)`dg!3L<{tV@)$O=zgwmYKH=W({IJ^z`kb<|ZZODg2lRE|P9LIDj3Y7cU-7C{kKLCH^E1WG8^oTsnUNaD)?U+SxrjH=891(+ye+f2{g^Dg|=wbxuDRM8PJyAk-EP!t*NkQz!#Q9RUKLBX(T6gx^At#6>9 zIi9H3HlZfk*P23Clv)kGNa*ZHC zZe_K2xxEw2WflvStx_R)I9OaRBE1Ffj=V}Z{8_Twxi-{R&T2@CGQ&Xb!nHLnE+33! zr42gA|F(C=Mp*HuvzpvsIOhiW1@ontx>Jn3$Q`H_8`H>57jI!b@)!-{;+1_fA1O373Mg#%X*{va`cr+OC1zva>EqrO7X0shKs)|fH!1e7 z05lIYFdg#@{rzfRP{nMJJAG%44VCx32dbo7q;ms@uUJFK2cAA%@_1+08K_X^d82X1 zgzyJ_p)7_FID(%Y&YaWBck)_L_Vbko6-m+Cn*%8>iJa~tIAA@JO}!y^${)A7FcjOUX0jY6bSbgh26#C+g^0hG*M2jmSbbXgOZ z4`Qc!0}Fg=rt+6N(R`W{JrE&UTGNkg<^d|{zUoYotkjqIs*y3kx1O2CYkbxkdsV7$ zfl74(w}n(`M&xocwIJP`Dw*@RvZH2|^0LSS-$3S3^Nh2=!&R!B{4%JzRXik6xy zD_RF4$j?GoWV60;qz+MFa&`a`x8^m_=Y!AJtmMU^5MO+Qx6Dkm2 zm5zlSU-`bNM@IHQu{ghF(f^vcX?&L$!u3q6-ySbM8)*%`SVk+JbK=Jt1jQ$$_s(G}jtC$Q%?K8gR3sPWU-?O-wQ)GS7wuCC0#@hP+~*f)gq z@eUs1Ch$vV^i9gGvY+rI`Yzh1Low8RteeHJf+?X0t(#jgpURnan2Am>%6eel!d14T!Sl2Xau^(t>Y)p_gATMu>*cDD&llX|^a zZjjO}3+d6O<9cU8dj;R1+#=vM6g-GLKEx+E&B#Ug#A34RHDmy3mg=?EdWrRFp3xRE zby7lj1mV#`sNCd*yjiHW;ZY0ip6>Ca4vXiM9*8bkO@J%iFUq08?0CFdm3(liI9;o< zn!!CUxH9sz7BDiP<};NA>IP!qDr6j>Eby#muOQneDzcHovQp4V)ka=;3O1zdD|J-W zC2Ufh9pMnKeLlLFz&Vu3!>Y(Gxtms1-oZkfZ`PPyk=HB=EvK;7mS~bvhuAh#dfSMM zfRHt?7ARd&YAW`luM!T8KBf2YUD@mzv*KTly$he2Z;$Q~h#t z3$*^f^}rp?SuI|4v-0w>Q*&AmJ$<`4jWUw_{WNj9hYmR{zjgShXLj*gskLa9w&ysC z-N%a3^wjplQ~9as0M>#{NE>0Vv6xnRt}JVQ7PttDc{Wuga3Op1do;C~iCco+e{WpK zAWuLD&;94iqjOqHwkV_PvSCR6;+EMJGD`z2Nlp&{8YV$%pwp{ zgm6OP|C||t@`3b_SI7y4{G#+Zhe#-j4&bOJDPCEoC^V(;AT-j#7)_XF#UrR|=L}_E z44$c7B5@RH?Mcod_bHm_sGLo)T|_Q1h^WgKp?RZAlJOdd$7%?M(v~S7Py+~G)msqm zo*^+2Gv*u=sPxm;cq8!C;7Z~Znc-OhM~XYhY#IKJhQW=Q%I_a=f(P8d)F39}o@5Jh zAu2kN2uGbFy*`RHTRwK?f<(D`Y3k{yCm{CEg*KGC6{FbO&xO8B#NMv^WleBuWM*&g z8-}>7Jv>}L9(?%No|?I`@uJO3SBSnX(5~1JeM4$kj)OjE-fGi`rI>L z_fF$_V#n%oK>RTLSH3*3@t^wQq4FN17Z1&7ufx0rp8eD_>v-YD8qO;toPvx1V+PkJ zbkC?2RM{&c%$E*G798kUg(ou3EB0yTbwi@U6$lB4Uar+)3xRQ_ilYu$;Ha_3IhzZO zbB!C6PHD92TXo0}oUMrD1C|NMI1|7D3x4kKI6)fbVr8|qiBtF@p1_v^l0du0whP+v7it%?&DN3F&-aRfSoO_juvMU+}ZL~}-q!dYZT;)rK9&QNQL zk)CMa1{^0xP^{Nszsa6{@%WfX*g$5Nr&GeJMGUER2}-*2LHAO`cT#%V98YD9Etpd3 zuy2Sg-g>exit9R2EUL5N>Z{XUWK3s)zgbz?x`OE6?Z~&t7ltw>ruA zI6_f^%=k#+^N36 zk`{tqU?p49LmKveSdyN0XRz-*WDtMwSb&;?#|^(alM!gfwcz3!0&Xr(?_93#WEPPf zV(D^uZffV%%hieL9q}&PUMf%MYxBizPC*&dLC!WIARBIz?}A_=)E6Ox^Unm#`A zAjAdMCI%VCYd>;9|FJYdzs1@-7I4mE&-8urmswBiR6m-+hP39DN_WG+b8Cn>=%wj| z-*`reKZdMr?BlfyKeXqXYfdA4bhG@LVrdTVqTP-nHN;WrYMd{;!2{WHDeZgA{pc7y=GoL=UrQMWuq>L3z3cqSkG%2(xQlcm;JSAqueKORh0?eW- zt$H2vg||17WJwby46(S$r@BrUq=Jii;dp*r)su$`MK{v2lDxi=&sbG8VD;otBxDRX zc)z^cCdSSnsXKZBOlGB2LsYWM^C=n#s1xDlNcf9T*H#_S2!K>-t&DJ7BGn8ftD$kC z-iSC?==L7=)S%hkqSRSP7TgpE4RzIMA#XC=1yFxf`4vFtlZm$UI-&w<9VS+#_sPYO z3(&pD8MVlm0y)29$!RP?nZXdik`qD@aH+Oxc@JdZ0$-`Mp%0Raf%5c{dLysN_&8hv zxMNs)<@baqO5!UQmx6?)%@|{|LcWUF>2gK5Qvgaz*b6VHT5D8LgQG;3$Vb9x(OVS7 zT&Clh{2b)ciKqjPYBpp!YHJ`lBoNVo?zYO;Z z>4tz%a1|cS;sELNS>e_}#0LQZj&2}P&V@F{6Ag%j14ti2=QcB1F4BI2`cG~Y#k!Fi zW>5FOT`h(;(C1)1zZE(pd<@&u<{0hYVHt%r{dD^48y@Qb+NamJ;ke$q&pji!C$x&~ zlBcu!so6gZ+i!-Do=9>VHIkcJ5SH8`{JvE?C0)5wO%UpJ!$Yn)WvGeXPkbkm?^^_s zps{ByqN&$f?V~B2FxG5S>24T!-fwt0-SPE*+F8I1-`6mZJU;lbYr!x~FT>3K{llL8 z77WApH4G%rrYC|}Fs0e+Fwi|Yq*P>qG;~+Pz;S=9bY{&oL1WKc+$*G;l&q(8Zu3B& zaUJsv{k_!J86vNfdL4H6^bjyDD0?>Xz?S&DIGJ34#dU0y2y4ys>ExY8O2uUO4IlN1 zA$y4}`pHFVToeF8=l0nvSX*#@U7th5OYK8J{2_qg%KQK*6{+8X`2$x(tr!aK8(1Z= z<|rR{sK*r))jffhX<4Z;9kERNZGCNXX^rf)&F_s+m<*eVYUKec-l}3v$7@x&mO~6| zcYEZ{ZsVd=SEV5oJ32U2TUikZ_6Fr1MK?S(;gtqWlfJ1JE3h|TM0)Kq3gV%*DUwj@ zWLkLN?Vd$nN#_ z%M|zC!V%8>^hu<^ZZPAgtKc6$=R_X7%~4^<^>5n5LCQZ8+FU;F?rC$@A+2;)-)z4t z(YpuM_-*@$(TvTub?@YsGKURJcdv!gER7I7og3}h<4V1ob9 zcXK=Z7MF8&Bk?S+Gb~>5aWwjw>HZSvV#r&&nIrl_$RN+$+AS0X&kK}Sh{N5$w;HVq zOu}SQq#6L_W(gKR)IDcUl2&!O3>)uihP+pS^%~q+$y`QcSy*n=Dy?R5LR{7L?l!}L zRMiR9S)e-$vKW^7f+281+uLDPCBw29sgPMO!rq2#ku)U)WO@71wp74UwtLF3iJFa> zJrGcW36Z7%QLeCh5rq^A2!#@SxXFo+=w(yX8bif}8dmcaL) z!zaU_+hTRY_k|t##RS~P)8>78p$1rW3x?tP>3jZ`m1wRv$s`~>3xpv%8VZJ?Hh>Mo za2*W;!qI*I`fd^Pj_7k2&n)wVm|MD1Dk0Q?jxEByJ)i;zRZtI|WdgCd0g(P2T1EgPpe0q9`DoS(mRgd)8+{dAs86r7A(W|)He)j zTfn|>Je7sFfJv#tVPLu~Oop1Tbh=)7#)+4smVZL?2(L1|2SUL-Tvp(OL*6TJK%p%B zx$fcxq*g+qHhBoNL6q}v<}*?tBh6JM-H_@1&E{k89IpcwWHho2mhmwMc}PvI2%i^p zD0Z(bg-7C?_PAdyUPMk4Or6zuGV5n*XpP*X^%{Jan&hPg#+Kw?{|&q^o4wRjnE7E5w0;|1i4d9 z;2QINX-@Qd+D{hR#cvNrqu1JE-zU=9-qxJb>TwvT4h_U@43{75d(VvHj{=J(eMdoO zUd7{6Mkp+QMzL{u^P*g9nctHb6)tdNxae)6uB=7tdIea%5TaZkO?lerbEek>tzqNa zdghTP5C2yo7!&dEis9i3e&*rVTA`~O@52%9nS4A!SiKv97{s#hU5F&>wk!m385V>7 zQ!Iu)5q5MK`#JF=oA{TcoqCDoh+I&gS!%2x1$fSVTdA$97IZPNq>TUKVQePH{M5sN zuwxv);${ql?pB-G(R0vX@JF$}Qrx1Haj5w8c(7O<&8bw6`sET*Gm3AH){1Lu2tOhh zYh2zI0{4p*yhh^(;9o2vZMBhK2w_3w|KuPCPg5CVbvwWxJ{~-npIJK?Jm}(_-(}k- z?c^AZeH;-|aJ7c$sRMm@!5$B$>E;+h+!|c$)y*H2l2^9MN=`{uI?}x^yNN5TdvrMW@3}4KsFT=k^+!ntCa=hBL2F z#!z|tpER2&3~ZE7bUJ@H7d$#(M$uCuU7<-O%IS5yR<&I&1(B|{$!TfRH@5C_sa-C$ z%cXkcJx;=ZuKyp}TJ#@eKjBHFX;EJI zZRyIPR140}Ra`^V7G*kXpb#3{ZH$>&ReyxUPjMTRN)1_MktS_?T#G;Wx?M&t6GT!e zQ&T?Zvg)o-l3jR5B(_|yqe`Z%=)xItVU}8>M7d;K8ewrPn41}4GnANRFn1VH;XSO_ z+N}x^Vrl5JuvIlpa{X3L}r+OYuwy4e)&kkPYSn1P}H?fpiR~nJuJx z;g~>?C=G>kz|UAHtF&|$Iwj?05sbcy48`S@F)4_U8Nb%BGv8^iymydvsV8jdWT6+( zVeZMgx1+8Ch9nrpHD#<3KlG-d3>ozDz(x@2!zF)v4_U^tkN>46d^HDFwBFT=5G^oz zxA8&b6i7&5)j8A`IMV#o3!MF=w!0h00>t;$2ZB>p@MgX5)XG>-%<@fM>Z+4G_t9HB^nBOil@~mjQ5z^x7mJ!lP3jB2$ z!~uw!tnD39m)!V3kHsW4@}x5|8H}z#&w5$-RQMbyFDyaZB1nSd!>>PZi^`8<5M5}R0+1w%=O#w2Cr=`*zX#DdbZsHRtJZ+8(CyXI+Y{8RKhkZk&^h&_} zP_h-bCY|&P&#YXTAyEj}?}U~rS{IN)*=L{C8%`~w)CDwYhX_{h`OdqHq7!Q40Vtk1 zT2t}u{e^!h_2=>N`V1Snq5SDdBCW?3!WkOD$XO)W6bylsBy}NxQstZMOz2!gx6aOXSF**7? zA3Ozr4kSBiG|J0R)_JOA+=uy>73OjiaSc_9M5q?mMc7#uc}WT_By1r*CSurSlR|yD z49~ARIs9I#)seA>@{nqMUbN+aheCye;pC)*EumYn9g*QRBRNV6gX%NLX3Cl{1};~D z!eFgf;2b~jU}LUHrhtM#tAqt=)d+lAM-212o|uuN!0_K15I!L97;`oLXaM96 z?FGP{J|edu@W1zf;A80uRZ2%ho1-c0A8UH4bk{dD7CS$Y&R(`aN~_0Vpn7sh)?$G( zbXUW`@%Z2@&H`bWUOcmR59Pg*m8uP|f4iIh*)%u(wK)Ub#wYf$FN9~Wb|rB%ZLA%D zP`f@(05fo>&ktiU8N)VORC;MnpcuqNFp(Uj!x`Kk4MTA&C&?!o{)Sb?gA(|tokZs zYOlkx&Y$zSC0y!-Y{ViPeWj&H6Xm`KGc)+@@J}qzCqg)Ocji00KvB!(dE+kLXF?q( zI--Bv$QWv3EMvS!p`r%@Ex=-%#sYmkt%ibA{BS}0u{1&3XD_=2hxw=rAea`e&d;TJcF@zfMGCh4uFql0IoJ{QMD_Z z+mK~zI+&0tyPag+csjikkZFtFi_h9%Pcm1xmDbyBxm>7UZtqYlV*|Uolh8?yQx-S0r${VGJUU?&d-RyH8(Q3TGzhyT&ERugCCW#F*iqX<;V4JSs>vUnJ z6+}DN$XMdftu52XN1(aC2vNJs(Y0(Puch>F7p2#a51?Ru{?i(FY7)%7_Yekj|0o4V z$yd538oxQL0nFmKn`0`oD99Dtz)o8qcyOl> z@P8R$FQGes<+mqq=(DHik0aQ1;Jg1&cYNEATmMUx;Rmy(I6EGDq%u7+LE(Sw5CEf6 zc=|Fz8{a?%Kt5ygGN4@-3Wds?!!yj;<L_qq}-Rn<=mLQRA`_=zj?{2j`%pH%V zLH1^Xopk#1ASjy9EOq$4=sl=$*@;VWHVDQTL2k&nz|(WOx~DV#kuBz>+R& zsc0v3L+G9#D0>ntR*GxQj|N@RRIrFTlBMQH&C&9fji)ioCsF}aI^>D;<~#O9-Xb1; zNwZ^S#u?@61&4HzL<#wdy4=1r7rxc~+#=%e8<4_BLYvMyajPdIjO$$o5xr3VI!wguiuo z65hY)LAepCI=`)|`AF1kkHHW_)#(rcEMtg~6GJOj`f=#B`tNr;+;{Dw}~>$OYGQVvBER!SMVA?Muw9GkpS zYgL;FWiB978>QoF6zKL0QXDE?)HW%H4dg%EeIp}d(Aa|PhP$~oN(^91%@uQ}qzck% z)Y>I#KNH$j?kSj21hXwF=V8y;k93c`Sf34bmFQGF_;>X3sHyO&_+X+%usBA-53p_3 zIO6N5x|D_K5&tNH&j803sb{|3Hm51;FK@2LSECk5D8s`PR<&^ zwTipcTJZC#DZnOzq`6m}okHZqv1AhiA-U2424u^s1Yh&n8 z@8?cjOV2FgIpBEAUVo6t-^bsurIj^TCWi(oCTy)?+Bodno*eZY$|~7FLSHJ<8BmqZnVp@>g;(jHhoI4ts!O8m&dx^* zABEQXkwEkj=0~;M>N`>v9)Ycv(^-rZ<`$Q0h%0Mv;vd|jkQ!z^&}!Ju2xmibO5Wi7pdw@B zaYseQaF#%$piD;M>{d>D90EoMjVfzq%F9umM|gm&R9dU#6h|jo0lA_1Y)~SA%-BIj zSDiFHQ=kL^RIhAgdt^e?iQ{bUICw;+G^$k!MWnS16@RN%X;+aA82Sv&PC4LNkBzus z4A0G4t-|^Y%86enHz&AKi#e!nVEkFkO+Elycp|jD91*@VmKf#A#Mj3cE+su3nn)s@ z=c(V&w4(rFoyX^*pc*J;#dkb<+T4R>_fe9C&7Ms>_>CUbI?z7ymQ$()(N!3H7+l?~ zx0nTk_0ZF|L+6pIYBNp3+pfZk%Aw|HO=#^VJJYd4#dV32$`lY4YN9ORg%)WenX za(=7LNr*}Htu$3`98V_RI;Uxf&y91oyt-a=o8dwY4%{m!{!D6v%!>1ZX6YxSA9LuV z@>Y>f2-TiyOD}E7hi+4GkSqxoYuWuNK3vWgQOgs6IQ>{Vpz?ClL<82h`^+G!<0L}}isVx0mFJ`In7=JRBY0qtskI#WM{Xi2oMH1qUFb{lm(F;rf0Guw(XNh7wyWH#UgRRQxSz}d!zeSKMVqm)$x5-xe`Q4WnSg%O>e=>~GD~FizCoF6iUrCPQP_DMxnvO$IWJpkza1nXLG^1HWFUj4R`rz=ZiQON zZA+Fw|SS_}2TtO2c5{dh zWZNB?7fDA@8X8wq!N|*en}v_XYjnlkAl(U#IdVq4MqC$F&jxr!7=s%{x|pf}5y+U$ z33ACPheD<5WY&Ux6oFQi)k}yBqkJm>7p7;Ro$|BPbf6e3r5Y`GQNr~(Ir*%VNiYbq zq_9Hz!K4uoTA@c6w>oAN#f`UaNi9kJtsE7C=}+i=);o!}r_K9^r9lhg;rr?v4#iWT zfQ}^D4AyLuwfD@(QL;LF<@Qph0{z?NpT8bj9+wH)!P}HSMnAq}w;dAoO(wy$QMOr35&ou*gqT9uTb4G)6#`MbI+>-0VP8~+Sv*D;C4pq$(9rdt6G3a< zcsnI%si(|)W(M;OOQ|-4PY`8LKcN}F0l-?q`=wJ-OUHo8H9bud-v_GYCa(;6GDmv% zJQYH1h%2|$`d-@#@rCC?A0|4^KEA1;QUmjG_AVv1b^)Ogixdb_wC!pG3Q@f&BY6BC zv8G}W>{7^?ObbMDI{WJ7>6sl>OY`W?*yQvKo*$W=&L28*XzIw}L)pWRj-i+)+odi~ z<)>FR;ruCUZ=NoGaA$Jm)fsaKmUg!qMNE6E!_(gDZ|GGrsIh48?sKF=kU9iOo7jTj zA&OLbz1H*GxJa~fV58X&3Guf6?9X{-<}TWS%;wS-?PcVEn0j>P(MJwF^2m|d?2)mS z+0c<~^fVcrA0Hl_pZlt3(iS@Rz%HwB$A9Y?Y`h^5^?Kik`uxqeD!I`tp^K=efv76e z@sa%DM-I~|j~qHOefW{=BV#Q?)~WTU!RH0|{^1wkOH&?XSPSrf z!smugxbrhp(~rn=j-s^K-G>Bi83wfg!b%-XOU;5c1E=n1kPO;T@5 zn8Kq=N8w$Bd)%cLU}{)8x^!rIX=-+AY6_O!)wMAyXCoyj_Vcm!AWRXf$Zmo<-I(<3 zq1i{pE-f2Hc&`N2F!PRngdO?cgb1xPtNp^xrjHy&^jD&5O+=k5(K}G}vmsRVKL1Kb zJ1?Y71#dt{EX}}i?W-?o+E$AvZi+KziFqHXx9Wt1U@$ufB^ksH^tp$~;?HUeNW;yn z4g5R$K!SZvn$`j~YqUvdFitzX&E<&je6!5J7Z?FLnNiCEfoKdiB+24|M`kW$d2QD*Inq+fuN)ksQw34w`a&|uQv zE|?!q6U=w}z*Ng^2JS?+^MK^r`eCrq=J6pYN~2|{UOaP;XDk-gn7!vRmN|#92^wnwO})GB3kK z5%~$dMEESQfgc)8ivp2U@>9plfFrnY!7fXO)RBETAbW##Wslo@N7SK0xNtpGj^<=} zfagF?G08@4f--nOk;#mYSISViV52|@U6Fcqd|cdb!F8_Ce2u<=w*=b4ra_~MoE$P9 zR=NT50+Kd#GBpsbHA%5ktn06a=*P$5J#D7@#<&H+03Um1>hOk*Ke9B8@a^nZLNF%U z^beX%ui)oRzrUMqEYN&!J#&x7vJs(6H@drB%|x>kr)t1EyNFk@63-jmamVu#Q+~5A zq{;NUy^uED{~j{)8{lIP%HtW9HF?9Z+~fZlas0nbc{K`4)AJLBrI~reC{51p;LlkO zUb_4;a*SQZ(`h`V$A`q@?3sy|IZ4^Pm4Ph14mG}5DA<=DS`(Q$a9g*1-VzEvE`oiJO+0+b(w@u78pF) zl@kgT?*=YxO*%Q_crf&3BF6F5QyOn-WM&)>_kkvhqxAI|=w6MfS%jOC8u(XCVGsT7 zlFg%O0-Wz}oy}A_{eI~I+s9$Us6m^r=}lW2Gf|B^BgZoui$V;ev3Axg=W6g!tdjtQ z1Y@*q;H!xra7vUC24x)+H5D50Fvn)$obPx@u9y_W5Ry%B0ovN!qG;^K<7)Y@@F)g- zS9l;S!H4=GTfU`z#%xlv1RKAPV$SL|32YPw%k0J|%>rr`b8IHnUE%C9^t4`H zZgD;UnvabYITbMj!_)6M=SQr?CwevkO;kfP;c)%V^|U%_Cx{=KVEF*O6#6_7>wnfz zl^UK{|H2H&#D$sBLknPVsj;|A3A~LgJ#-g=NYpXNR_fjc@u!nT#24fyh_Y$ZMbaOD zN8+N_`m&_IjAo$cTn37wD{BaxQ+$xD(!@xWj4j>!&1n`W79nb$XnrM+P1PMZRYANH z_oZ<7Um?f-GIEQo6)%Ayh8P*TLn11Q3lv(uW{BuX>{@shp)WFAfEQ!NbOGxa9r#)o zsjhI#$}CXLMY4DXt%;MGVj(bH=E#K=&X&d9f=gngdJ$1);1a4d3=Pa@fy?x=O@hPZgw6^j4&-SI?kA>x1PkLay06B7Nxb&e20YEgTW3Vb6}Knf7uCOD>eI%BB? z{)J3YTPbr?ftJl*D6}eNXkf@Y!+T+?Uaqc`Rdo{_&9q(km_|8t=sbLolXxKZ$Ar9b z73vfkP}CoWqGUK!SdxSgQtaeSf=ZCdvIq4dKiA#xEeY?u2{C}(al`Dj1BM55;lkQy zk}zd((shX0XKfHOQkV(B4YsR{4=B|g3Ml7MN}{uY*z6HLW$LdA|04? zH%N4e3W^*)Ow0z~)DauZ9D!+;SVsoQrYTNgt&WgqVgRuGtU|e9^N!r{)S*y26mtl?d|1EnoTp|SbLF)MNzSD4S93`4LurZH+VhkIl+hEmr$xQ7D{>0k;|6T^| zV?(KYbW|srCPk+;Vkp9M%m*3lEUP_jbShViGSSQqAYQ9@CG_B!6v&T(nFGZ+zzO_iF7@wNgWgYMJtNtxOx7vbaKGzeILfu83O)7lQCXCI#d$pt*ASTKa=o4u3{d z7c*=$>#Y^!v$3&kc*IQq&|6z)kgu_9&nMXNo;FXXgmnvm)Vk~&DvJwjUo|Sw8>(~l z=HS6xJg@nXyP@f77(%2@FTHRd)CrebD7P3+4D4zV2PLxVH<~rQG2lS7LP5!N!^T^m zwl&8-@s8sB_7U3}S2{2uZqiMW5F}I%Or#rY(aFuK1$Hi{LqyaCr>9(NRY(P|)u}K& z_B!1TQt*fR_%!ZvD7o1wvqe3@3W~)9p&yr!*bl` z=(`>2J6fHT?nZC7+scNA7D&qN!)jJ6eg3k<+Q}M+;*Ie^JaFnWLz<|J9 zq7jiq5k)js>An`%tgEV7TrS4O1(CPb$#RGu@Z3#<$oUVKkV;BwFOq4&m{G1SA-DvqwC7v(Mx-_@Z=x8>5&@81E3Yq= zSLGKneW3zwqgJk(QFHLe;Lbd1ue?@m82hY)&NG(MNw17+K(wzclVZ z@f^%=Ro4$Ma7)a{Gogt^pro>(A*Enb!duAaAF8$3m9MsZv0Q64 zC|l#j8p0Ad%^<7i*s^$vax4Mjla;_fK5owW@p01mH_;fCw2GWu1W1qkVxmS=xw5A#e(qbW~JXYG-M!uJI>CtuJM|Jg?`0dTFV1z3Sd7hd2h9m{94J|5U3zl zZ&gHo>t?%B%p-~_JK8A1uee@??1Y#3`pOH1dSSD1esm4qn)$T?5|y?`Ux9dxhV}B= zs5se9VBMRo2vY6i@&5aNunIYILOo_;*mG%Ty}aY_sH0on^9Zj zJ*O1e+hg)m%C1eHixlAJ%da4%H~&`*$$#=#n4lOFk3_fXQ!$sUn$!8I(0jsBv0~m& z%l9;_%n4qpt;|gsNpTmwgj@WNP?R2e^_FHy;dj8n29gy1g=T#NNDB4;PP4$@lKq4y z(Qg)KwU@n;!iPXh&%-!hg@-q`QE`i018~-AQ#>g!WRo+gh`n)*NU&Tl%8LfphBry` zS4XaP2j_ry>1@owk7S_K&c`Yv3}uI5En40#a?)fZUL^5kd|Vh7C?GO0=l3c(l4I%+ z8p@4Kp~8i;z+T8WnQc|U9U3fjh;RkE&?rC_5mr&!iAmyA4ayAU8_fsa9&W3luO>Zr-mfl${glpx6=`(&`Zh#h+_>?VSzT(`KsQ z^|of;|F@obI-Verk(akjx!1GN!iaRD`uOkaov5H^VdS}{@hE@O8 z4!yf`lelYd1l9oNnWtwLA5pn(9!e7*x4Jvknr-+^?3|-=3O$zQocBs*Vog>*^3BqT zZ@p14ASDi~pZfIZs3+*@*}i5i#o%6)cW-ZwX8{t`$^%)!6#a5=(|I| z3n`q26GtewX9gm(D^Sc0?k+ig3j^uFY_EmF>ZZtV2egn#;P*}tz6gYm0xCtKxRVSovn7&q zvIo??MH@LMySRk_7&c=51GjeC8nLjryG{ZtS7@CK+u23+72Xnb&3Z9(^tpB=eMX9O zeCns+r0NNJ`i4+X(c+stM8DODZ)zqiz8w}m@GSZFu}RIu2JFCtk$Bc_0byACAQqVt z{74TTJSep0tMK0p2cMB@2_>m&d`nro6k3=FX=4SE6X+IoVqu6qL;Wu^4}&ocf~#2^ zK1lcb2p42+Kvx4F=W}Cv=1@SjE}-VY{!9gO=ni_=kuhzyFyY^4a1c?F37)|bYy1?( z=!qQ~99HvSnp->`=!4sU$}!Qk%HwE*B7L}xMJOW9$zX~MTI9APe)b%q_X`?aFm!5z z!3)#5@@%erSpMJicCIj+D;&zsMwvbK=D67;gHTVL9U6Nsst6hRwu_!L~aq>5dRE5bqRQE#~ zTyzOLV{SHgnExOB1z>a20094;&K<#z8TAJq=BD}AGzt+8@jn_|%0(P6&ko@Rv=Y-H zB%Cs$PGRa?$3L@2NT_4`3Dv(If+90uGEy;G*yTC9JcqXZ38^gX@*JVh?DCu|z;m>G z#NB#R@FY^T-M=3SBPpCLMEMkCsWOuY zz*_Z8E||>)hgjdn`9t}0Ns(H=6EPKlP7cYL74mOp!vbO{RcaZdod?IreGVoM3VaX&1x2lWcTS{(Nuy+@ z)+02BpVMU7+eGhabChs9V=Hi=otxcYM$29tMr_lgBxj|QGG z2iNO=FGSUe^!m@{G?P^n&wBk`{_U;Lg-(wK9*E-ga%&R9>E#cyPxV<9_vBo%3bsyS zx)A3xr@@24@}Q=z=E<(Pw03DNE8Ik#c;Qm0LR=31hN+4|M);~d>f`_ zRHEjfc(MsdH6CrtsI!if07ZvJ9~dPSW~)MJdvL-K<3oWT`0ZB2Lnw1KbVkpQtD{RC zizuvH9c{v3OPmrOCM$XPxEE1wh;Dt?5X=ULGkBwMDqegJjS`9=iWfKXazW0Isvg-* z1&hKk_tIwSe;Py1PICbnt}!Xf zj_ezI5=Ey_{D_eDG-@7iF|_8;1DaGKHE|~7f_c>sh)Rx4)snZC+gXqXjemkX@d#`6 z+9iF9&uxHs)3;7mu!+*)o2^YZVzAvkOkoR-gV{=9b9uE8G#?L|=bL$j#<@B3H#}Rn zvtzGl-@AWX~CkSULV*frn$MVr!&EVQT+ByHtDtw z*RN2lLx3VW|2X7L(gfhy>!;7tEol;kzT&QqTvkk+n zh9feJi}TQjoIsZ32@>WI3w{xrfR+}g2Q6*@8EHvNfGgvx8f5Du=3Kl+t-xYF%){AR_(YsA798nI;|6`hPyqgItv?PiW zEm)%4#ljQi-2BBf;dyk>tz-Dhz+HH-5YLJ%0yC^=057M`rFm_~IuHBRE18$A-!J3$ z8SD2M{Fata>#4GZ;$5{^UnlOOOk=hY50NUkVUpH8Wt_Tt5+ng+lbSHaw+O(A$y02U8e1C5+;t1uNE8Sbt+;4Q|~s#8bSTyew!s|*Xt$nU^bz} zqFFGqxQqhzNc)O^E-&nipr&aPnbA;J8!^u)pP_Q}jBG7Z6f`bWam>m3MW#5aIS=^; zU<*=4P=FExLr?X->x5pQ`#VD2yIaZLkCwjs)nC(_NKKE=+J}-6poO}mU-%H=)M~ob zC2EC<@MDb%WBzTR4i7b6pVJ7Sbb02QKWZOC$^~vs(8WFVo1XYrmsH`NOjZ_H=F|| zSja3u(yVUfUn|yYjck^(5X_I|4kJgw!uiK@QBO}C_>lcr%dBt5sp)m)o%)@isjwtzAD=zN**7#X4j|6H`ET?|{7KnQcoJ!f zvz-@SB2N#MsSU4+GviC z@=`s>F6b?vN?#hSyhr~P);qjC67A8iXhIoakM4g$6U>)oKjBHFx<^0x0`d|9vEr)p zegyYOu1?pYY!RTaSJbzHY6mAIx~ik?h{Yh;XBCRvy0Fc_^Jfk5vXs~gSGOwBKH3RA zkCM%Zi&GB8rSEmk#7o2`x;Oc)?$oT^9)1Hl{xgkq=}x+)6PoXR-*P>ed?KUvyZ-I2 z_qDUG?$mpKcW$J+=R)(1(nVs_&~-{hX9{ZFA;Ev_s-sgw`8Jq~gC((0@eJo+QE!Q% z7|y~)_Dm!lpYo&zGEb`TU2|NJx_?OH*um3XA3h#Dn4eiY7(6J_t2)V7vx<9Bde3GX zau50b!J`lCZdF`8P=z~A&q~p1*iq!zKndX_i=$&O?tgM_qaP$s534_#*C^(t62SUq z!|&q@ajUrmFG}r&2)Da> zZ5w`yI2;JerCTc%(A#p{Mwi951CeruZk1LkH}wu#K)M zFBAMwsJ{l>7z^8Z$h}K4Q&W9(gzU6ukV5t45jYvjw=xYD-X33C1j8uRH!qW?=FVbn zQvHVKzbWV4nj4+pEK{_9@b#G6BNc z>ORyI*Kwbt?-QZET^h0p(Yi=@>L)ezsJXE&je>*mi(#QDS^HQF-0H5b(k{Y}4Hm+0 zFv81e;^Er+EgqUmJ58ljRbS8tvVE-% z@80$L()677Wft)iOREqIOLK?EM9Z1jmXFGov#~9Y$d=RHCJCO^OpoWY-On?b_7df! z5YSj=3_PDv&u84{vvb^f)_$sF#N$eS!?*3FSxCMwC^Z`#hpI8K3?}o$3W1EJ!Fgj? zfoNT;)ar8wml1}9)c-<)JV*#QKgiAaOC*Nc}U~0$j8lcvr^0+R3sOKC|I_N zl!57!h9d||`AjSzFv>0$u#l5dlnt~VGy(LEjgQZaX~}taJMC_#$nChhof2&)eH^nG z{2I>RumoyLA3g73F6o1N05SONztDnAQ}z>{M5^nvFLBx7yU4Ul~}j~hQ2RaC_^y0lT3I}`GDXB0KhaPE{R zDuuR4Iz{#xIKT#3#N%+@B}l^*1w$Db>XeibWzcK&3u8gX4ykFIyPoT5_LgKLenYtV z4!s6G0d=ngSIK4ix2AxOeuU1xFVtBgiFEne{nv1UTQeD!NS{j)!xrSJb{mVpkkH)6cunhr*-+0D|m)I>L98O~YOGxZa zaWSz&m*fsr0{ZDwFPs$@b&$)Q_Mu{A2t~vWUulc-W2(}a+?JF|y-`Kgv4thBa0Ur> zA@d>qw=$e(s!OHvDrz-j7A^Lp=Ya4`ajUjcB0p9{hePb8f|0Q5qn-?W{wVT~{UO2F zikTuRX~CLHZ6Rk>YU>SfzhZXca4vuZ3pJpEcJrZ@}O`mJx;C6w>=%gRIdoOB3v6gySQ@Js5ZqPLXT~ z&+2K2@=DZf12Y2XZaXtS(U8oItJNxIV=sGFt_d^wB`Q9F6b#{&0)WUoSiZ;+PIIHc zJgT{sS3)-Busu^)w`m%LFi~A|iEQ~x5 zBq-CutQ2ZmoZle#5qzX`Ik;$x;Aysc-d#@`3*n_<$BrgdYqfxv+YYNJsjL76Av) zk?L73T7Oyg6Q0EII+78tI(L|g@)K}HnIJ#jHwvZTl{W}h`?dB|!Rryce&`jdVOPqy z1?`^;gf~fR*&1Y8c-c&lRjp8;Aj6(;b(>(ZVuJ4)ChVIAYYOsXkDnZWEH5fTABQIR z#jo5X6_(OIS)Kb95t$&EHy_oGK%G|LM1sEzczJ7 zDBYkDxKU7x%aE!N3B~J)V(p!muF$BK=6c130C_zcA}=TQwzJ=kl@XvTLiWAVq|9f> z`$Jv#rd+z`lQA=Y&ybLsT2B&E4U&_P)w>~~jo(Kyd9`yTBi#7Iu4fkKnLS->mTD-z zfz+R+S{Ytkq*JX$k)OEz1-%UPtw!u%pJSb0?PoDcwB5edti*Y}-J;j<4}diH%LV;q zG)8~+X~g4`vnU+2s7fs5CWpsYy|&td8VL&`Y?LdMrw&HL(}k51wlSE#&=6r+ED-mX z)k((3TSy{SM@ApiQbFz`c-5^I)}eqRybC_8LYD$A1|S5uV_Z_YA@nOhMZ!I0J$6|&Z`G|Yldk%vcbSAuno4f436v7JA{{Dnph78; z=MFglNay36_6$e1a)_~OP(;YBo86z;lud#2n^0BMetnBa)thnSO+x5_jepzEa*dD@IhOThQZ#+?IHiB&6bVA=DoSl1G)T9&l2Vt;PN!v z?!iO6jBEk$Tz}MIeVLRx?q)%T7(YofT8URn)9bay%f?4(iU&4c>u0I>VEv3)r%U~T zX??^x>%^`Kcq*j~p386h5>Q7)KCw zSSqiVkVUGFdJCIK>$F;2&1YULZn=gDsZ8ZANCyMu_@$jc((kpF+WP+8hO8Bg%zoaZ)~v^3Ta%7$>cijNooh3u4>tX( z@s(v3cF$WZk}|E`$G+apoyuKaLR4#VkM)*4#w4LyWA*9)^@!3~5l#8AD{_~AYiQPn z+AjXR(@njR%!j$1MHW8Jv+%!x_N7l38Cv;1kxt~*A~{a!5Lr5u>*Xyr#^SIRIop(k z51xg*a?8lljvnx~Ak(wSC>fP!H5h3l;Tu>PWfFkD6g?-auSxgLAHx@hOtbnca7NVR z1_MIRjTR!U)EgvLwJ#{PcY?X|+e>h`*+ziE;wpl5>Cem6iRqnJUOk7kEv- zx(~uMCB73m^dYApn1q;MwMoWbvet}`BRL*8SP{|-rXn6ez1f1ANYB)4kogWZSp{Om zWYwFld5(lerLq@tRNY-p4JM<(F|_JWy@6?|Qowa)t=`t}%*|Bkty)!Ui@kMYf>(Sw zv}+w3)y5A55D#P`u(p{Z^QYenec77>TUS%xZ}_VknfdFPbV=GwwUc(waX9Nd+{boO zjJL0EG>hY9_;+b8gYS}At&?>Qb)8k7HuuM}32TuOH1^>=3%0A}jhprMS6H>4XqO&yv~A4J&B$CUdbNS)b(#Oas3FL1!+$ECdCS^iJND!Q@=^E zaW!}Iz)apeLw|?Toc)8EYe$_O7HA1S@WI_X2j=#Ug5Jr2iN_iTyj!2?3v_9>??rFt zujC1+=iRxY!+6jfFOFFV@rq-2fG^BjTAmQdf=wuws*!j7d`rdi=i6j0pi_Js7}d(Nhd|q*@Qh=`mz;7Boc{zG$fNK`P&>B7&(TVhv@{1UI=Xj zNABo6m}m{umP^=LQ#9|yjnD)Vi7)@yh%X8N7GEAF&YzGPEEM+jSuD~V&eQfYBs_M} zu#+YlZgnZ!4Alw0=?|Es-1QS_0&|ylBU&)^Xb=lpLUNUIcf!{0s{Z3N5%IM#TtyGv ziEHUWO+2%=_MBn%{rj6eFoJCAlmi`#WEpPw6jG845cDf`Tqw9BfB*h|ozSoP#O!2o<11Nr7Ej;|9#zQn0 zmd+&Bmp`w+jP{NDz*%W)2(ryWeZAO3@R+t1;E0yu_m~kTN7_qec~ZyF3l7xjzPq

csU5cIm8bEqG4YkR$i^2(^K7WGC!!2$jiBDDay%W!w2n-TkSoLpI)zC4BJ$HL z7ZEJa{#$%ULwz9KX%o&}WCGQ7njAvjh)UN!Gsv?4h@s-V{GO@!eZ%0{qW$6e!H!Y0 z@-zAwR}V?DqKXd;@~OMej-`ok#U%6)9rYv@2|d|De{AEWG{MOx-53_FB>yH3B^%f6 zH_+Nld#2o~^t98>R9V7<=;)5>R#9Ip~@yY~G+SA-VygLPBe}p;{@l8e}w) z(irkzj$ybOu{<4Npurb-LgpWM5Q|2Z z=Nw8yA#%;!|2&^EimZ*3de8ub zK|bhj>PgK3zmVqCf9P)p;To3qH)O(|rH|xqd>?Pomyfx1zHk&99bVtK^rmPic#2c8 zH42BKO9I?}lU%)0LImRq1u#-t4R#bF8v(oA(v5n(n1NF{W$s@UrXVg&Bs24aEt-D8 zgyhXKMN1No4Bse26vv3@64_Ic`JO7y)r(+C4a^F+<|x5d09&HEaws4$!l|v7*SK69 zADU#e4dP(3wpDyvp#96C?cp$sNWuH}LSH802|sRlf*P55!kxZ_uwe1OkDY&14Xb~b z=KS~B6l2ZXeeRhr*0lQ~H8eA{c`ijFw5FG8w;t${Q`5f32GHJnQPYEoHkpNogdcby zEcIO><7*Q`+M_~-+BlpuZR|OO|2t^zWDMs}!nXZL9}X;YPb8j=<-pMk?_SS=5HZK| zb*pzDUtjpO>x9rMS*6V=Nl>+YkVGY|jA)5;)AOtPf)0WwmmiUiRwP;~rjJev7YlIP zq@Y&{XfG8hCDbNK$c`jS*LYd>%!KDvVYzk@84UR1PMJfxo0l6IxRxTFl;HjlesQvp zgo++$4&&+~T^k`WZIVB6Z;A0aF|RJ{6q3%lB!*{56OxZp#VtgNF}Z3fPM#hjGz;AD zH(P>nIkT2Ev;`AVZn4m4;u;h+C`%Cit2L^1gG2BF%2L4(4DPn@Vk4*|_A(>rUS=SMK)KzgeB$Xdl67sI6#nVGF!bIb zxQ{ixp}O&4-iIf(vdZ{~Ha98$XHCUC^YrTy=0Hm4w}h5-C<1*?45PN-ifwvj!2PZ? z!QF%XXo18(J=2a}eU)$ut?r9Fw5*Y08hYli;q3i)cAbAphVi`7qD;s05<~q=Unq^) z!29fT*n-%7?g20{k`G3i-o)W z)(CJH{N6Emtm%qHb%S@t^{2LFNTm-y3N{&-IExM75od29Ea*_&?HgXbw z<3Y|KG<1`aVHLW}0~a-S9vMKT?X^wbO9Iw_0<(p!p}H9ufk5+3N@%wxW1e|>cInm; zAd+s}sWcbhhTc}|u)Gv8^`stoN7J18dt!7TYe{08dR8jgCPJ(0NfWXL6Tdw@x+1|n zJ+w2m|9ZaH+p*KL2?k&43)L|~!d|H{T2SgoK0qC3@<(muybm=~A!6X~HKO>uYxISX znB)CstpZypu^#%6;s}IySwfM~oE6AD-fDn4JsM!xZ|)1sn05DAzQSxFQhnqb4U75v zEMJ`uo>JBjZA=ix3|u(qz&^gTl^K&h(h^BnbiyQqK{9F}N6XU2h%rK(O0$1!z0>U% zYxjMjwR1V{UJ02Z68&3~1=IPDggQy2U;ctIfT&5Z0mMfDs)5p@;|aj(-Ae$!agy%M z=}HSWGL!d#O-$j%^bkoHbuQrlme=cU-E z2d;SDW|6y>x9z|J4r@8Mh%o-+R)nHysJLjYRasT3hLQP7nADkriMpaFP+hCEtF_HC z{C^y`AZnffk4`rS@o}C}VElgo)>fyrD^gRAEGym%4 znaSB5{5gw1C#J>UhxqTETt>+QgMF;QW8vW;U@OXfuX7{r!0LsqH@+2rwKkWF^*7E? z9W=T{eGrf!E#2!Eos{h1G~5gK!Z{DrUn*RrR15fzInLckGhNYL92}_0s6FYLoQ+#O z7w8$`XPQxuC{)S&Gl4D9sF6pa2N)iNlsU3BHt&E#fthj6cUqE*M?UuQpVHBXW# z;LXC>C`be-BE_(ocRn~t$*2ldy6eQ=Mu`9t63XkymLdFJ$jAZTm#swvKH`!Ny~!8O zwI1&JjNoucjolB9tPMBV_aN*?LL0^v7yjsBEV9g+FmHxq;(0RkWg>Cmv)`-9MUBkj z!d>3QHX|Yjit?+pse$s_s(#eL<3P~a3re8vq_&Oz?e2uHOB0w|UD~vu2)~J){!NOm z)%?u^7l73?0f;aYMPvI$vG3Gm)lYY-m(rZ-UXiIkZy{Mf@?h6@9)v+c8{AlHhGF&W zJofbNlHM&Ar%@Pjp1ElG_B7FQx1Y0Fu=HvX2UcVl;SZ+?!~D=p*aA0&&ODfjCMR|E z=#s_F`dm;%VLfq4;^)&u=S}Jqvan*Fd4O}LJ&^JGpQQSGJbAHP zcyRVuxzt*g{$H${kfjA`ToVQ68e(9mAtTkf*(v-Fp2yboO0^1AS`aam7UsB^0W~YVPH~m842(kt}@YCT;Zxu z1sa1l?=DDZZVov1aL=WbHA*KMw)lTeL<>$TKP> zWICH)rz`wnBdRMzv#1^j7n0uG!9sj|ThDU1=gwZqYZTJ;Z;e;hMs9tttN8aundJi9 z3&e|8hp$z z#W8uuU#P4Xm+J+2iw*&R+}>d+QiWdhEvLA@vMS#Yi|J@kEH40D-GPQw=OD_KW$t?hYU;XL=4Oqe-(x z44tdUfUdL*BH^mZx2n>uS~PLmWebX_3w*jf0upT+{ckKIfq zr&6G~4_yF0nkE1$g@_)aqn^Z0{GDCwRL$P=M)%L#@!Qe_;r1{Sw^pM2T`X98a0Dno z_-}XWpG|Y>FJHN%#zMlC93>W=*VB=w@F{+{2>t0a5qdYf{&t_OddN<%79K$21v~46 z@CtS(aV_~`nW|RRmdUn7?mRFsDd}gchU-X5K-xw;U=wLUP~55v16d_U_8gepa0vz9 zB{a7_!V^e%5VVxs1ObpD^j9E@AyN_g)Io8WzITES(q$#t5q10_W=vC%UUTevKzEjUmas$WYR8Tf z0w%Lc^+jNDBauwXYPil6=+PXOiHO_|zVmI0Qfte-ZKnt_|KCF!&Jo8VGuZ!D=*vV> z)Q=e{N{!4?)F#AV{U5aRW2hENFQC46)C^EVYM`FI#+lUjX0*HJKs07 zKhNiWSvzPL#F^`{&S4cyuCc;V@N_;Q3jpj4I8EquaU9r;3Voya+}hCU36^;T)3=La zI{XhAA@q=HGue+^OiYBjO2iO9|0f!DH5F!v5Ab;!1EC(jlc>1|+<2o5>-Z0uil~Nwm95@itfry+xT{l)x{Bba4IB{yVPWBt-`{ z%oel`#5<1_QH4Y5kHJOwT-55i+N}Sjp5w-A#Lo_FZ1=;~ILPX&M?AA~7(-_kTmc># zjrJz~fha@CUewh>&|aERCDGIc!>i|v)H2Yw8Z1WrHxG~cH$UtF(n9?ng#D-kga6hu z*!Z+S?058y*z5kDXG|BdPdkWZ_R7eW3^ykkE;8yc=hcW7KbeJ~{+vN}YW->Oc>#WK z_yySiPdvy-wg5jP7T~A)UVwvN_KfK+zz5w0kYb@&10^F)<+Qkn;t2Y0vlyyj_j+8Z z?%n12{P4>&^PfGavXh;) z2*`^2M8;(VmQix|d;otRM5t5LC>&lY1>9B9C*%@CxtX`%nG9!Zx-i<7S2M**k#a-x z!xP{qr~#i*PCI0nYc$~%ExsUoB^!gp0BjN(SzM_#E@wB$nmLBFkFX_;EH+?P$ZllB zBTR{IV^kHj3ilMYre_ad*0qyO1Kax*8VQ!Nm>Ud2xS9nLmaK!l#Uz|VnHB&Cz!U)ip;c~r|z`> zZkp46VmRh!0X%F^9#kdX5*81(W@JD0^zIVBHqA@^p?|x`|Nb-P}7~1h5#M5dj>TdEt6^4=bhCXPPI~b&AAfX`sJ6HLZTZLVtN`2EQmLvumQi zJVhFgu}Oz5^!Rex>2=!ab-G8d({8U}Lp^$(aeAF`dY$Rf>x|oLFMEA?M_cPL5URRa zNI`8k`HPTJ?lDl=9jHt;dcs}Q#Q?g)S2<@idO$-6)17S5V@zu64$F#q&-;O0dbqqx z54fH)oR>v~p9QBJ@{;29KhoNcVMdm=`ttbrnHr1?DjAKC_qXQY%SBO@!kx@yufUzb zW)U85EA#>iE&k7>P9`;m6cGx0xPN?HN#>;+HS1-lem%thH?q))#=h~5oj010l%}S{l(vTq zjfPtIr)S#nYhn#x*2I;D$mP~CB6>~!IB_8p3A_LFjxhVr;J162^KAg1y*TIb2?QSg z1fmXq0wKrui8!vsCy-|N6UZ_A2_zWbCo^&suhBo~%j&Pi4IY<-2RXEmIeeol34^2= z@;K;5V6QL5m_*f~CJ2$`GOYpYk)5 zd3fPqC@j2BYFX^bA>p%*r=hz7O+Oc+Y4=3F%elo;js;5Y>+#klnRB555}nUq{ZUOM zYJz+|->VbcbuFWXm#9{;An$iTW^z#Q`d+!fuBQp?JSTSSRi09h9bj zyGVL1O(flhB44nP|)Yf!VTZO1x$Fdr)AO%J8^V{OB`sza{4AUd!~ioE|hgVj1_R!Tp9* zbKd+7Exs6LWMtfh8dU#zB!;7!(d7NWZVwzvi!^X;Jm4%K$S~4|i^8V(MTd;zw-{u? zr_CZ1qRXO|=kn#VJLI9KQtl$VKaL?0|Bqpfy%Tvcas60FHEj4;kQ)XkZL1Uz0ZYa8 z)@yLaDs7fmAd9S{cqhCm#%SK0-;GSAg*9p^jY}C4R;1Y!`By^eeS1%Di>Ajg6O>jG zyLu6|RN0qY_=l|2Dy>b_B<0-8Nbmv$fxb}5i78R>a#8u7Q~a8`XE#1W^Ft%vI-FDJ z45!_%gw9*v?Pa))d(4m9LMwdFA_zeVv;BPHZjdj9rkhASxbDyOmQbJw@t~V7E&Os_ z1eaQ;A8TXPJ*@r>dwOk$GT z(&?%D`NXBtYo8pzF5HVskEg<(EZn8kVI0aMW0>zu6GQLn$$0g$bb7b(z!5LM7TQ8%bh~KWzO##76ly(h3}!tqG0wC7Aohqt zY~N%Vi@2j{(5(SBmF_&t5YK}w%JA`^&U;^?8w*PAEWDLrwiN!zB(~D_BMPFtHk-8u zCAOhV*pT@t8MfG%IDv=iLnCQpo!0)qMlmr7z?+pXF>W%{Z(!VL?>FuX(3{{pyYcni zkA7yKX|?b}d#}0Xf#7Iu69E7La`eDSx{(q8l1RUWKT-a z|G&L=0gvlC&IBn+q{J(Ul4whI1X;cUA)^}t8V{0sP=+m#0416vsDO|Jo1}Ug-3@d% z8{Msb5P%pY8D*Tcvz<-iz?;`>GULfQvzbhil^stKXXo4aOX6`dnaQWblg#GZIG^Lp zW50RDn|yi9dS?Hs^SrljKkua_Cy*1c`}VnYs_N9KQ>RXys!Fk3FXmShm2LXOKNsr` z=bJqH{5Fn`N(odfLxA|U?1KuJ(N_hdTyvbG?n|_Lp$$)sO0m#ztmW4m;6CpvIECa| zzL>2!&o5p4z{2U}B{G>xCt!TAz4H?W^e{Tm!BW6KJ zVL{4QYcca%ySF9M0q?3`90z+oFc|sDlSDua+)L&&@{MY?PC*8WxB}8gnO?_<)6`bL zAXWQ~WfJuP#qEIQ;6Z8xH(2Az}m^+-Pl-7rV43^SVeRdSzMuAek zM8l5ZELXCnm_cs%0@mFHoaW3-O+TK%^WpgTXqDc8Fp)tZcA{HILPgosDMLHGEVi}J zvnV?R0s=M*flgrSoJp>2uEj)$H)rjo6bk6f#zq%M z$KBBn(Vsc`vmh=y@JCsUZ?xrIWZbXk@bWY&{hZ+c(Wm36{WE2W*y0qnwmXW^qkG$7HK?KLi=b zie(r(cjpuw;4S-Jr#Dx(V-OAM{|uRh#5K4>_&A+Da~m+;z}DWEjnr0tqp{&F z>bTw74b^GbjulfIs~MbdoOJ7#>j}li*;(^B0f4di)wnaKQM?S)t~wvGzJCbc7p?D$ z_%4OodaA@6zsX9I_K!;NuM+zf;rqeLO6gl98@;}>6=wzSmMV!?v(<7f7As<|&&9`2 z#N*D&<&)!6BnQ*{ahg8P@y8kdI5rlWNlZ;1OH7Hj$JEn_Y5VbU`!Nl~at>E|N_mK7 z1um^RS?8kAoml6bd9ImtL6pp56`@Bbwi5UJoZxB3c-e2ExCOb#NWokloz~i6?&TON zKE?U_%P}gT#-FDrX%ytz7_;H1EQosx_TGX`w$0vxy|-ZRE!aWE!rp?-7VTviqTYfX|Now=pf%?i56Oqzf_62X*y^`PGYNEDXc`_VP;VN32YMCi1ws$ zgztfUaTF+jfRV9$e)0i`__${3VlKYeSoI!?B{G4D0xTOKfZrQ zl(rpyR*$bDiM^nU5wldI+Kd`0dvVcpK{UT zYQp+Nd~itx2%~mh>lh>U*yxG*DA2OhOf>K?0$=ovfLBXB;2}#hLTRIRKkIR3x%>lP z7d>g6c{9!mMbbSy|F|gRvsa02^gW`Qr zuRURs%}|O+vL_S_FCU#}LMyYSU+SnTRvyAZF?k5#e=Jww*JQt;a+|5ZejRYmaZPkX z9l`xp4w+iHcwI#v360~ceg(a<3O_A}c}a!SQW`;&WJfhs#A893ii;0$V<9984rTtSBT{-MseZkF~|n zw*od@hv`=2uvQBMfA$W~<53RJxd`Wg#pg)Dc4gR|*5p%+^E~r1@NAR}yhBfR-AU6t z3qZa5T4PP6c5BUVJEvnF%YyIQ#AmOpRa5C3(>pgxJEFz_DmbZxlXN#M2GN?ZxFt#! z#9yBAyh$yu+_=Io*u8A;OxRyFoE-&bx&L>>t5mT2IJ}fe5u0M!jW?zdZ=L zHUEQu$84zj=APGD9_ZH14Ijv!maEsMHTx7wyj)H3G7p2MXO`%ZMJ|82%=i3m07JT~ z2McvG@?KN)ANQahdYt4z2|0V8?j7D%6W`?V{075X4?Iul-SAWTV=G!F&A?hy`Z>;_ zqS7MLUEVShQ=lAaEe``3)x==6diLZdcoGR2agyFBrRzuusXYzR0o1%`Mh6p<k{0~DNUbGZn$7Z1DQ$C7d*H>Ko4FGd&_BMAmNi9HJ0FC^`X|29{r@@w?dosXq%0nf3a_@i z)J@yX^lJ?nW8|j5_VWD?!fTG#KiOI8$q+#N1ZpNh+*2IM%mRg=wmO`6C`pJF9?^W8 zc^nMD7g>Q%lI8_9QeGmx?N$4DY;ze1ECyHuls|R`g_Tq2s0hu}s0%eiKK{jMyEk$h zIb_VA|Kytn-uZKY-Op+c$#35A{Bs|Gp61c#ep8R2LV)Mdy?!QJ#JB#|75!Ug6zcfo zMcbM4x(-62eYaN-bWcs+4*>{wH-Ic&8I@qIlXop!dR)=#eW(#CJW6B_6^`~pJ+Cys zVn)nh;Z;oENCu5jdEa1I>!IXDY#QSja%c6%pX#Y-2G$aNrjvK1p%61%aJ(2NZGI^a35N3SVQwbr57-;tt3&8mpml zDm6jOS-Jsh4@S#s7Ik28UxpMmzb&H=i}`EWVm?vB=eaZ*66axpm4K7sF zjcgs-^JpY^-jMQ6^hd7IDt*6+X*lPbhF;~{#WD5)-^)E|urCk2Q|+Tw=d5@W{SBE- zcAswskR4n?Fop)E^a}>GYh=S*DOEF7HhXkw>aHD<^H9Q`82e2G0BP(dNza z@Fw)E8O$pF+rq2(ub$GA-Z0px_?3r{a2dujTw)^u42CvuRoUmV*$Q;3%eM)6dr4^n z9Gv(TabJalsyTT;9Vb#0y04WkX>aSGDS;4d+^fJ56kC1{S>wnskac63EG|}L5Q0y6 z&tYG~TbGSv#i-^}wEY(V37Y(kO!AZ?!|`#e0wp2j6$=88f?E$1x)eY}Ev;YKi$xTo zIz+jIv#?aj@y74SHLqT6$OE;6E4Ie~Nj1vhioj};W`ANS4{8yC4c6rPGMy&a^KW;3R$Rh zXcW-1cObf06`zih(ZkBxv*gS-<1MOsX&pdbXU3`kMYM`*$?m3HETsB)DFC;=SpQp5 zvi?4{lWpB-5f$)b0P;OE_DnQd*FrEVh+(Ljk75z9Xo?127>dqBvm#kEMFlYoRi|mu z3b{a9Bt-%)3_W*Ri?xN9=Wzi1zEsVa@`fjPh9h=x;{r{w8_KQ9%yX5UogF)h=dE{4|JRc^EkP{FFGG?;x{;53 zpWbU3F)==d=3~%h6+~-^y)GQV8>o+>ZrCUtP(}LNd^M|PVPEfPiV78)mqUe8Khu-X zIEH^dfCjIqf5j2fdIjr4J&#>`pc>h^p??>?slpB}Tl?M&XvFbkmu16lVQn6#7b4YE zH7s&^$*)}b2~MvGP8BL2aQ~;itmPmjcAd znhN)-Fzi~Rh%7((l9Kn%$KOm$uTXzJgiucish;5&Ow!&E z-&$h>u2>z4NwQYh5wVgSkQu4u2FUGS8+!EcvuOTsc+p&%2oMHKG~WeRX`Z72If7Ef zL{jb>82L(0yR_0+H@h^1lK#oQ<#}~EfCR5R|4{$(Y_8w2BHbZd|AF`-Si_SG^H1V3y1seu;{rmEd>g8?{Qv{EszWh2J2V87h3C zZ_#<}69JffV~CDfl;5hN0}S;LvFQA6c+vUT7Xk*%5}o};eq+H@B{PuJ05H`Dy)Hdfe+P0 zXh?7{zL#qhoR3q+Jm3un*wWo2U?JGfqRe{ch*(lx zoZi!&oX(6^`;i;8H|a2qws*{G`gQ6}^Ok4PD8@GAp81We9%;i!GmBR4x^>?^ERK}? zUxaXTYJL%2O8!=Ft}@IYabFa%P(exceV!MU^PiLI(=3Dckuo3qZ_W58svF?M!PO`v z9CQIH;i^+;ak-VMosx@B1Orf(;gO1ag@Ob|#*ssj-NxKzq%wPAgafj~?O3$LYlr}# z#B3b*C6xoEC3b5RGT@AykR8*TY_OWbt0}h%Hxo|$aVk$?@3nlr!3XEaLIK}xsq9h>P0f!to zwFUwauIFoLQj}SuF`BYsLJjc%*I@q?ksv3(MtKhDY@g{Zme)Bn05K8kIJx9FI6wwC z%807Ul5L3m(=}up+{)!w;Zm`%&)4%rVJ@G}Ibay#GE{<+MhzJ`DG&mgJ}LUycL`Zh zS8JqmBu!&u=U*5bgM3kpLm9jzEP>3}*d+`XGE7(97gI%qN6gzT@J8&?i}P{r6=g*; zVrVHa5V;N^80eoQPmG=8$%)Xaf=3d&AAje*TlUd^}d zE-I~Q>Dx|1wZQcAidP{$<3pvV#pcURt%4LOyx1HTf!InXSnTru(v=BMcT(dkJ010j z;z}qOUN&8dl1;DEm1 znm6bFwJ&zc#D(pI>z0J!QQ^HLh^=;icdv;BJ!rGv{37;Jw5!EcDvALwm&V3Uy7Ogs zyWGGg374;UtAs$>jSAg_GVrmkc!` zmz&(ys=b-mjJe(103+{5X$$#Aev*|1K9gLU49bv_)d zi#Du_;b2)Qhzg&lv?=z~>{s~D*@!`tw8HAHz5c9=_JZbp%IGLd^A^^HX8CVDD5;<} zPua+$i=l$QRd?j~|9<4ES1HW<7ce;1?j%Sh1+@uKmZq7l9{dU^xK+g90@S7Hx4rn8 zHZ}zvf!PgvBh}Kqg?}hy%ki-x6w3x6JrS$8#dpo<#2^@KBJbnppVapDAQsPO$g z=&GPLQwCCDzbK>PZ`K|8Bm0Ls#(@oB{wf{(lReNC)Fwb_nr598Q*f(@!3C&I({Fn{ zXWG~lbOdHM^y}bYZ~;2lH0&3*03AFyumBxAIIsX6JUFlb9XvR&03AFyumBxAIIsX6 zJUFlb9XvR&03GZH)?JO=AEd^-QNJ1goq{Wf&$@AiYPQ<(Ept2kXvcH8%Umv)sl6-$~L*_FLqzv+&FkH!2YWZTh zgg{_XL(t{Y&?*a#UI7n#VVFg%>Hh_5(rMMczWfIIg@RU-bs@ssZxdit$~4FMIM}8m zGoiJR2H- zds8LV1CpAW^RtILBWbON-C9Dce9|Yh-t{r4c$qxE4f@GDUO(dl-P6kWv6Fg)6l%OO z=2gPBo6l!WMf2;}X~0l)U%@SA1uicyo~NqzHBQHf9K4Mx=l-XABke*&_)NE!M}P?~ zN1((iOQpESfy4}a%53U3-YoEX01CfOA~aO^XZ=VLFMS_3Vid@kpBLF17ob4*x(L`&;s(zU0hnAlXhCS%d7PiVrNK8cvA%#V_)isggb}Ck5#$? zL&`fau`aKyBHWb=58#Vf`0I(QD7;Mo7f44^<|SWr=T&QVq$nEiAM4OhQdgH*GbJQu5R zwz+630I#^ua7W7YeCr^jM~V%z2cPZiH)vZ#{2h2vY|fRd(mPK(77_nMN?QtVmvH=- zF(WM^{(-q7ba)AC;B{2G9OtM^N?M?51tYE!(N#69k}&ynRh?8ZU*Dz>%dxS=#ztiu z7vkiNbyDK*3kTDhkYOQVsy(hpLwI_<0htFFeI<(Wo0fy7`gy4Eejga`klNLysUp0L zx_td3dN36lta{s2)lqGf4(YBHtPa9yg$rxAU};bKVy#|#CAKo|stG>sR?}tZh}lKh zs`1NHSLKW@Rm6Lw9ROq;nj)_F88K8nE<8Vz7!m)}u*ae8{Jf%va)9~yxu4NP`aucD zk3^*N^YJR-3@uM0`e23Xy=LE-8sR_Y2tp$*rc64qdJ2N%?xWJ?XSjD|osBU_8bYnCS z_?diN04`?Nac>>WsHLuFlVDYBSRhP&Up#+lIPNgla^=mK(gAy;_l-GvS^==G{T|GI zznbew*EY4<7(my4`D-`YH6ys%GoQRn@!3DSVmYZO9(<20qoKdriu~@^9dpAxEhnrU(44g&Qk&fSW$B4 zjiNpxX+^oZ?beEUB*aKKxVJ^c-=O8tCvs=Vy09%UWNie#up$es(}GdnED59)7PiSY z-Vf=Pf!5!9SV#Vm76@sMd2U(kU($um=%ZH%Kka+9;}TSmXHs7sd6)8ys;=1tn+rX5 z2D@eiz=BxgxpsJ$Eh%bw_de9@YI>go_UHTJoitzD@$R8cVrsFabGrbFgErP`)Ra+- zHt%RN6RfV=HVNp}yCd+i4qIyU=zZ9M&tHj_&mE?+MeIZ3P7E{y?c97*o8(lkP6Pfh65yvC;E90O9UCAxl<6S)0vQz(03s1{K!Hj>d*KxdD%HF@BQm zF6nQcb|=U|f_$Jc*$@v@OK5d>ChktQw!kZN${}Bg$3;suncsw-^$_qX|BmMdT;<=6 z7YFuu>_4wH^nDV}Orqe1{6hH2J;mjgHhb z&Ecf=!AcdjLJab%_vvC+6jug#vr`jI3XlJ@pu^Alo~*oi&42-{yx||#yqJ@4{76J< z<-Nyi>ZCF_^MH9YHg@6C0%oo-=`3&71HtVr_0_HgL?f*^6Xbom6pG{sLt zg`y8YXQM|h+%vzdS*j3Vqvw#J&h%tG3_!TM=CwKtC(E0pdBG8-dX+vEnztm}Of?jy zKm(zZZI$d*#}lf_*w{>bvg9~oo)Gl56Xi|oUylXP!X`%CKLL80g&)-5^h^h4NQuV?WYE&sLtuwrQQ(Xv4P^3Tvf8aGMK0~Y06-V zn#x;LzLNYiIQ}FJ)Ugbv(^3Z0SJ$$)GMG+F8BFfFpDJHYYZ**Qs1zV^I_Z zv@p~>+*$Won09L!fJI;G=1i38W(ekPyF!B&O&yyB5Y$;MSyT9kUYX6BU`&IcUjusU zY%Pq#TClV&SqtPbg0`K|hSYbRmB$|Hj3r=M9Jx}SZ&U8rYB1yIPv=pn0H&g)Qz^d|*0z2c%8-EoRYEXz-$Ue;`LK;@UI_ASX!jR9|XA^pf0V zb^l0|guXpUJ}f$ee-9wjOM)`iWhlA&>K(BkiIR|aYD?ZH%oZ{2+jJ%9R!j*cf&AG^ z%1=j0%0V&pEX38505m&`m!;S|#Z#JGRi$WubND(U? zpS5d2`pSsO@_BqiiOYy_q~wk5on&4WjX)JLr4nAA+3^6Q*RI@Xtw7QC`kI2dtWmOW z*<86q>J^>Hx@Puo-9QS+S9bJesXeab>T!!J-ud%em_L3`h3zhB9>m&JO;vV3icgJ= z)ogWlk%HQ2d;chM6JkYu?eFOAy%9`nMNMw<2_mh3K_U5KutkZQ6v7u$>r@&-HRVt# zYrTbyB7tPZB?5V8f|kkqS~{|3a=?$6)PLwxA6{kpHe2u{o=O)5{^sTFzxM&}X<_`> zKhoS($nnBBM7zN@{ISUF*sLqTW{H>Uhk4*7?f;9Cw6~Mx*)poS!)EaB0mO7yq84`p zWp(pxm!isne_+IvCDc*I3Ck0{9sOOvIFWVm~!X~UQ9ZNa1aKj3vq zcqP2Ys?-nrfb=BEA2*V$kYGuEv?tbCGTq|~U5S0L{nDsrb*Cnp@oe=@N6C`Ay~3b- zYWjW%AhEm5TjOA4-eWj4|EeYb0y~m}clos(7V;>6_I^i zT4}6%DEYh9_w0D%*#Hu}v*SkxoE`q#x}Ta3(j@kI ztD~vMk)`i{2rqqKE(HvorDne%Hbk&gxk8|DM&ibEo1NGk#>Zy!%Fd+yRd3XIPv*1N z-H{X*B1eFX&~TH=SAdj5h_awQWlnX^h3r+jkWD!n)OOAtSxx2v1**`1Fc}&i=}MAi zIQGyO3(61pri|A@p0Qcb8;6uJImoMC8TorY)OxbJzxr`46^b;ryB`pY_79xJuXZiE zGU=V|24AtFeFHDSpNf*;d;OAUk=puOS0X(Z$5(e0Y5vDvf_^PZg6{E66^ovhpSqHB zq;;Mtl7fEnlJ&2nBrB){jFMdCsPDebvBW-bDfX8gX82CGqoLicCVDGV zZhlSl37;@{4GT{(pH}pgF7DRM%jw5`U3Qc|6@WA|qj{@PW`m)pBImmQo@=73K6+Q} z7J_D0dgn;5Rmxi#Ew>SzC)8`}d5d$e4p0+4WN(_ctlln$pGBuWs7Kl`(o%2lKD$QF zSs3e7xJ+FC6iHNs)u@EBYdej3Aw?32;xGLOCM(9JZx6uLJLn6Z41Uj%B1HLfoJD#Z zmz@W}Xu^n~Y0`PnHo_0u>Q4kP!Wy(~wZG$Gqm7oSXoa)`2&>`h=BhZkH=!1La7SD& zZs*?_;BK&IulAfwA3Kuct=@>Z-9x7qPYsNMS;E5rfq?D!1k9^!3L`d$I_EqFCp?>P0b_1_(Kq)7Qm2351JaW^{n$f~C`(@A8MQm zhtBUKJM(t>r>_n8>xBR;yaW2N-~knuW>5N9C}X+RkR~d>SoACoMY%X^=*0mqZs2K^ za;TCN(GpVFc|@!Aw<2nKP*_mz$B2Z~?4SF-fFZI3Wo?;^-h7pk1lF?^k%@wCZk4S| zR$?EcZpOJI+w$NcCElDUS2M^XTPYW}ONhx%7002=k*S(;NMpr7-PSBGrDX0N8=yZ+ zTMzU)WcyK{Yf#oS8alI*672gy7hi4jqzr>BcRpyFCohijQ5;>wK-MUlJ~G^GN*vEU+pSw8rOobbAqX$}#Mm^P6GAS%k88OL{Br} zs|SabNup5TnQ)(&2^POBiUNM@N`l;$ijrhdH2S-jh-#EXV9SQ{SqpLbpjTw4)eb`< zJ8eL*e1AAfWZ%_U@-&;8T^a@u*IlbvVqm73SF8NQfRPepB~H=yZbDdtm!&m0y>p|~ zx?v2C7(&1VftjXVqqS*77`?JP*oyE30DZ&Yt4Ee)>PI0{A#IBN2|cIHU|Kr!R?G?| z+5A~!Q21Gr=%K<-_aoTitmlP*KD`3;zl9K>X%--{jlg9M2o4DfM2Rg;v-7|*CQVR( z&{^*yWY)oNAdK}8vJm}Jcp)18mVg1Zgy@53D2s@T9Og)^$k}pZ?aZb)nVFe#X|(RD zvILwP9T}lyZ2TjSxUMswMZOX{VEp<1yZG|A|4@fcjQVaa_7A%g`s8!_=e|jkThli$ zZ-4J<)t!8w-rDHz-eLQq4?_DDym=5lZ#cDgbJyjSm0Z??f^|vj4$WBIThl!Ayd|#o zcV8}M{y~6ncxU^xb7eiXu|cWsTV|=4$5;TA_YaEPZ2AXyQxnG0|DVH4|EXW^O5hWc zz$9feZbM@0dS%EQ_I8__lG7^Xuh#6Bh+b<d?EZ$ed5zySw&$! zygO)R)wX1nW$1*uBzTD834u9gwmyqcv{5{(jkE}1Olc7QdQ z;Z1iL1#YcJBJZ}-C|$#yTT0%YU0chi^XQ$+xzh0EM^Ni-j!M^1(JPIo@O8>^DU_po zy-|e0`C_Wr$hz}cN*Z3Sn~zuRWrt!t>3tJEMRGowMQ7IdG`7WJn5V^Zt+nXb}<`RNqus zChh%!f@ERWra?fTUPqgB^yHP^*WbM}_9H&_w>Q+STydY4)5!Hi=MC&Xd7HdyOFvTo%BLNqNp7-=P-Sq%BdwnAStL{4|O9+gqXc z9_xvj33Z;co65P!l0JNOxA`{w>I#RpPECTJNAUA|uI@tCa*j7^pvkqF>yTmb1WJ(b zStO437i7Uk@||DekOJ&9p=e8Rg)cVJ2nDF3j0)*!HAe<~TQj`yf6{CkD*V$f%OH_Q zNfEzb=Fw;Q6>FXQRjhNTvcMSUT1eMkE`h8@7L{E@NlO#>w~YrW6nQ}Y21GNYsBUAz zeMx%z%e1O-(KasR=%~7vs+cA5Mub~5YSdI5=AwKTEsPtf()Pq^IfHYFR56Xkiq!~~{&J^(sD>I77n2KQN(#c>DP#0_*Z`80#VI}6X4y?EW z4NBUJVx5dF)^YhS&i$}x)497S8S7joAwpVav1oFXNHv>flpSy`ovmGs@|={O$}Fdc*3QX z)2Zh-HeZ-}G2($y{iex5$YASQuK)L^Jy=f}(;rI912AgYuECIzijR54LY= z5!a~|5dPC1rbs{=@_Ir)S^cOd0Wi&~s@6}nQ07iIPe-8Pc)^{)iu4<1MN;#cExUt4 z1={G!CxegK1I>8P8rAB!kCy^)^OkVA6?$4N{`^@lSRAqnnbG`vL9N~B>oH1TC3Uhe2I&?^8%T5pWnu|l6J0Gr#a~pRx;FD z#(G^PQ(=<=YPQP_v6|PgIHS-j*&)b?A3Wu1%PHhN8A!g}qg!G+Nv7XnBChr3UEi^Yr2oc`OT;h^)nJ(L^~} zf)+s|HWs7r@y&A!nUn>GH+~EJbMeaun6iDBlA<7eCdtKuV06wTxo+3E_!UKU%Ed-KD}HuRA^IgYiLsne z>0&l^QN9>C!|ApV&SGGPvc+C=o?p87frZn{$)$@Imy`2nmvCk`yu^1QYQxT%v(LV; zw2(YIZ#|tn!+_4fDTEgDbIWt`_jv&2r<22ulV9VCt>s#x0{b|T-O8iBdd%OMEHa2z z8`t2@YW=+(AdQ#>A$5&hVzrq0t=-!a>410DFAjWrYZ!pkdNxS}pwjz#GN17lMO4HU z;GQ(Sj)Sq>R=^-u3lR9!8HDV7oQ8*53IMwHvCFu_BVaIfAanIi9OqT5scqrlgjbaPr2Gn1E?>`QYYst0)$vTCihhy+8xy#VyCrPP zYO}FX@^QnxIy&x-auKNQ(RiF*LsfJFvW=qX44__4$}%}I&#&X6Vohy!6vvxzQ-Cbd zDgGuh)pDhjLTztVUkb21ZzKFoU@f03Y_5VeF56*Q+^W)+B$=k*k2)hDTa#)Flu?9- z?s?#Td!^>ApcB+>do^1vqY73LBOsy1$B%=ymCGl`j}cMw>DXB8cw%buSYk^2He>z9 zji;>#Xgp&#o}%BTSr-UPnJ~1iw$Yk8XqM&_#ScO`RP969auwvF)VWI+!<=++4M3r% z66ZP0a7mxk#^MPSn2)WDyNaK2M-*nn3wSL7CvzsBMhi8BFwA{b?TM+R>|CkrWWgb; z`g_1fSuBK{{;mtr{y!c7!v zqj9Ckf#f(3G3Gu^vr0Y5`h$kkf1t0f^Z3joVc$SfrV`Um((?^{B!xr@!DEQARFSI8 z%a%c+r(V&ow+bLPks`>SApvia2y*vO0B;o$H24GvOD7rY@>>)lsBqIFD8$1OMESrY zj#wvQ(Nbh6ubholGT>@Z|EaA+En81+;g~O5j8#WRu2gg7%U59R<*FMu);4gRe&X29 z)e$Tnvgx*JDpnyE8+z7`bn+%z5p1y(=|uy2(O$U7x+V#pN<4wH(9PMIaW3tt*vCE0 zjzSvYe!!6)uISS0+7&wM)K=JTdUkYUG_VV(_O4KwksDL^n84TOHAQxf)v5Nw1fXc28B0v#X#h@M8+1445* zQbo7|ve)w*rGW5&`X+@Pqlu%#cF9{yeq0h9u?@G5%Mmz{Lw+Z?EQA2U6VGARvd%DG zLPro!ihDp40M{oQl7V0L%7=rMk&Qj}LcC_Zpr z9HF_(RfPZ0MFyJ3%cyA1?mwk<`+G?bdEuLB-V7Bk;vfIg=Tpmkt>gbOgM6M}F=}e% zll%AWJIe{nN_4(hB$a`!fwcT{-=v+tmCg9M43>xE}gD)x!v08LP z120kNmp-C{9J0Qn5MZcS2b!-`#sUo7(!dV4EAkaea9njM1Jy>shdI zWM=Z%2=Qy=_~eNZ(Ow?3f*#c)W5rsWibOiNep#=S!Fa0RSxYxWN@ng06f)=eZjoH% z;ER=j5lu%+4>2m^aJZ`Pb(k{c2{@yRAYg24HJ_?YsDm=F80ciO<+Zhh`y4G;n^~ub zaNIhD+|{zh>tf)m*-92`5UowZI8y3?aFDZGP6&Am$XP}9R64t~Fqb9~_zxmFN{+Fb z79x>XOVvjN5Hs>>4+~mAo6{_5n0XyJ(0T4y^h5%#Z)&VX>PYK6w$KauHJ~?dyxLv% zfw$71-iKO0HoeaQd&@U2olRL8qx&C>+Eio4rtC;BI>@58xAp-X2r_aNCe~Kvf##WL z6rP21R1m{Z_1-Z0*PcAg?=10$PN#KGUdSOkicUYdz*C&};{M=@1Odr95J6!?l1Z86yp4^&I+CEc00JK5R-uTVgud?g+iEjc#(!leb(rGg`PU&2X0b%i^Z zXdrOm;3`B?H7FZ{2aO5UWNd6EK3Te10n!<Csh12lpTB>%10%t3;1FVq$VQL7>nn_R5&xW)r_Hp}Eh9{>72)@UmjU(hNoI zRpE?@u^te_i|B6*p1(O}u5K@8vG0X#i34a99QAihJZZV@CQRC3$g_puA@6&>^MPdB zfS|*8Q6&0RrwCpTeaC>HuWss)lRG6G|EXxd0YPf7Z@lcrRnQ#Q=T{Rr5t|^(aUx%u zVA}_$qc|$hOpqbSznC*=>pkHo!+S6c&EX%LaAf5WM7-xa0XL_(;&{6AdGxz+66&79 z0U|sBIOXJIdIVRc&Kc zh2l=O8QeG{Prv_nVzy;d^Wh>K(D$dW4gBR3dwj(PS zJ~{N529mCT(iuE`m))Ymq3*X9$6Zh)NdXgZc54nzj{tw4JXnsxpkFc6XXJF!@i~3h zUAX9j;GNcXutJ-<3kC(H8Q@Nx)5AEZm{_z&04#`A0cnE`Eg~Kdl$;z~TrCp9KncLz z>yh^zrix;5-iKO8w4S^cdlmo+@s#fP+l=Z^JmZts+kE+H_EQ?005R>zdrOYyqnO^- zI79<46h%?!8k?d>;Dw>+@c|W|k;y?J1Ym8ffOZE4mgv%|*OCHni1e`1+ml9Z)5@aO zY#Y5MLo)63vk$Q?BKRzy~oSmBRv^FK3-c=PD#rLz}Ky3Zp7 zHJz&DVOylDl#v#B^pNEb5mSiKQY7`%R+H8v=%K>T^-cR-hW6L>__SY@NGW{e&;@XBy(8r@@dDZ9Trb7E?0 zN+wT$uw?6sjnM%#yp9eggz-1~9+1OkKz?1bwWk5G-t2Wio;bTW|H38zm{4g3oGBne z+Fd2%3h|GK@1qnubR*mr>d~dxckR_bN61u-7ojYUF*$WxqI&Pb8}$s>0GKx zw($~dTSTWfHa51kG3>^W%EZOV8+J=bz_D6Jg^L;zm7wL*?v%?}!#8#La~MdG!Zf20 zDOilm*ZLlrBW7g&s=tx3-t2W`9$&n;a?)MQQ#vZQyaf$J8Qjy*Ln1dSHW0`PBjkO0 z%0=$gBJ`EIw%;7|b>jEOzw`t9_`jXqSnj~Upupq%zgeuOYB$nRN$^7VcEjT%@@2ng zz}|vb%HDdjS02Zfuo%ukkdbN%;p`i$5Pnkl)i@Vb+_yGVJNOXBYgc7$!ruI#Eui8~^Jk%#U&>NFcyJZLH^KVqu?ef`#ARHwzy&Ec~wi zSZKZ3D+@EHPP!w+o!F^s2+WyEJbooSP8gdGq-_y@ALr)k4YU|^=#^5FwSm8Z6 z=v~9$fnEnk|EM3CVe@v_NcJE0$5HFeUOD>2g@ySSPr4V7{1GZaA5i!Vwj6Hv@wH

P^T~J#qdhuh?0qjnowT>-GnLEeDL|z^<1A#s{9WP8AXPhTf=xhnhs~e%7N|IJ(5{ z%SZ4}mrWTTC%(rF$HV-J?TMryzdClXv_grvXLUp-Jp8ztpv*JC2FlI3RNDL?RDo4BjRIkq?z?=`Z zcd+~p7p0?H4y&iwgISgb`76?AB7E1^0ucdp=bJBbAp8IgP-YSF2=mdDep6}}VF;-8SVjI`AX-=W290Ie|d_D5yzhr+i@IDRA| z)e847Q>^wTB_6n5M!@#-nBpm|iQD|D8CCdk!f~kZ2~Dz?wH;>`n~bNP)^vuG{tpX@ zg)AfSJE^fVRU~_ldF41kvjQolQPNt&Ar+cX^~(>%pOB=ik6h!&2MB9v?%Z_ zdH{0Z7vP>|(c82D4KRy7_TRKD-7Devk%)8_&8X}+o0NuVtpUwdhP_$7=E`h6>2kH2 zEs7*d;(D4=)q}jMSzI_y)D0D8G_!iu;-7f*J<9e9HQQDWW6uIbUf!snf)Xq2N-AGP zr3wVi+EkO0CzUSk66}kuUz766-mJ9AQI4Qmg^sUMs6vb{^-&NmO)!`$^^H`;%0op} z%GKtYLV9V+QcDZIyb{P)_K>6r2GnvDM;+9fF5Yo-z4B135I+%xRCp50Ub*DYtOZyi zj^zF&z?~(iYE9bX4Bf)?az+a->xMIsTK}}{NjIF<;y!?Gc>TPV{AVN_KN6AZhCe++ z0!ji2{|}CR(`6)wl&XMb6&n&-fl*`^t)MT)#>oE0t!`{cvA~o|>xjiyGKZ67NbT4d zyV}u)3_*Ndx-d4zIm`%(sD&v&;eb@O8XLpG8-2!+6od6Z#d3s1VL&BQE{!5mUxerb zAs~TpRLW z`OQ$KejX}(Ta=_76yj%5p`LU_c>mG`g)MzFsPYd(h3}6N&$k6k4Xew*pSx0UE2fqr zfj=7>#7{$opNNu(1I%m-ZTX-(qL}v@QTl8l%0CSiJ{u*XxgJyu?i56zWDFsQ{rLK* zwN?ew-TP^2hH$pl!?3kzvmmxDR8^6-Ewizisd1MQJvdrL&$^dKStU^UzoiSKR~@-A zQ5KeR^57j9W_k7(k)qFXEkFd5#}q7?io`1X6q`oXB~;(jE!=T8|JZb#^Mp{5D7n+Y zdT^DWZR45fpT9DwG$*Z%!9dtX8O~JoTDCenT1GV`sviK#x8v_jUNS2xMJd^dL zgwIObUVEJ4f?SkjP8zpQBwr4^cKs*CAy2qQ^PI(b?7 z8-yM}c?9)<^Z~v?ILJecm$j~7RS|bSXy&l=1uMpUUHa0@Wy!pAc|gg$121|{2)%Y( z#}mu1{yRN72PB-EDwz}T!?mQ$fqkuIlni}NheHQmYB>6QOEb0*Cz^%|qnbs%O6~r8 zN2Bo@EMqP69Om%^>>-sJHu20T;q`9i54kPb0f^P?$*D)(lkUvaqqOsfgBb;9t6*e# zBT4Z{>;OoW=%NGy9}~;8*Fj=+uJepzz8V~N*;0Bt(VC*@gDOQ4KG5ay;1oq4M1%%@ zO|ln~s&meu^t9Fe>5Pn4k4bcHHvNG?b#Tvd7=-&Ec<}M68{j z#tiZ+hMv~VC-u}DVBI|Ogr0>bBpg2yk*=Ey%ktz@dEr{J{A+(4&JPef%rm=fY{iLz zz*+m20Rcecy`*1I6P%7MlVlLEgeZ*WCz@hAm zmDg|eX=*X}A~~SSczNDe(c1Cq_AFh3(s3GGCRvuL7FCj=7-y1PE!si)hmwRn>-xds zkXrvNN@iP{m@=yIsSPhBZkE)#aJkbSN3RrEKpjVBaZAd}z|i0Zu6(A`z&X_oIG<p&?xU~l&B&0c5_O!2+hi!k~$)JzY0_D!F?UMurkVOjVMJlmdV8=rSWn zh7WKkjB^r^RF#T$A&M=Z-BdoBP%!jh7rsDz9x8mk%XnjF!8bRSep`>wukb4-I)3F> z;Ci`0evg(KV~8h2@-$qXrwCaTox=GqE}Xz+0aB|ls43O4NK-(xo8tzuO|vLiW{hma zHt}`mgu4ifu3W-lbEZs}pUKH0(k06T$WC^vj+Aj!u&aiF#c@6#?B!fYHIVcG} zy=EtplcpM5G?*Er(t+QnjQn@l#UcUS8C=$El(?K25Tln<#fi;w6$Q)`uNdKc%@NR% z5)T)Uai4gdgulUDqW{EcB&J4CQEh6 zACSPtaWI&q*&7?4bWf2q+z`iL`i!SJXUpO(h zhzScR3>Bxs&o+yaD4>O*WGfT%kriv$}2EoO&9laY7J%~0{k1RwMr_Ty&B8a>y_Hc$;oUfv6;VyYP7gT zpD0(?C+XYd95+XOCbW&SG03Ne=mVB5A0k%DFuTNE6iy-I>5kt+Tp8g=B&?-i8E@01 zMycIg85T1N(38=LQnL0;#|a9Vp5-C~=6=}oh+Fi z0aI;3jshG6=;|mAFTe=69Bx>1Bn4` z?odK$K{d|;;I^N8xzo+D-UWM?_JcD>a*#4pRSf}0r{3d}d<(&rpMn?~V9+c0CXYL~ zdM26~WuX@p#4uF7KjZ|r$O#Q93>}AkipDG){$B!mX)miJeB)l>BWHx)Oj`MNtP`iQ z*y~ZXr3G^~A9qhNk4Exc`XobXM8e^V9C1>Rt2FQledU zN!I4$<%9y+@#1w}Ze#)vb{mj~(4o5_P~Wi6;G=+pB#rCk_?6Q*0Sk&Th3dvSfvr1d zk*5e9Z0~C@tjLg}#Y3aA$hF7Gcg`wz~uoKs!COoV#l;DbWUmC=$8vVW*fMxepC2vfc zz*Xh){(bxWAxK=Mq(;jM_7|$k6KVK7v?c@)3kS9#7(t<**?G$=%QzMlDWdp^t~){9 z!y(M-79|B~n=gJ~yB=SmWw9K{osEOklnr=w;p4_lu53-zHo?WMOYo9s;MXSAo~`0~ zmO_ev`>}UabRypq)4oyZEj5YlNxj$Vjf|L16s;nb5IDh|SF=@kk|~%JD}vZ>!AC9a zm>6+}E}KzgFk!&c5!ut%OOvt}M_kp6GX;1}En`&vgK!O;@9mEUM=&{4xizQiAa0zhs4vmDt-K2xVk*N{p+sO0?R+XGcx`hpc zKReDp^DS523uxgY%>Jbhc25yz6*Es1ZfyD=A7nCE%4x9oG1V0oolXS;+VBb-avoaf zM*^;IYODqQfnc3rq1diPK(AgYJM2{}>L}9x*eGoOFX-XVqLj6}J$lU|(f31F%EHZ` zLoa$AhhgSV`+0ATam#6!k!!2fsqlK;Vrdl60>&Z8$r?z_Ghvq)iI6XJ!%cO z%k$8z0N&?-eI-gt4hj?7qC-6iprU6@lkn)xt`?ltd%tf6VChM)UD$sFD~q@LA#yR*5Mks(Z=#_!-o%B`6O?dRw95;dfXYo|1nYog>|fZXdaN}a;xFj zeUE2xA#jiPS$Ga%i^IYf9t2#6rcD%WO5nKTkR;kWIDC}j*2p);`x4p}$PJBq9sWUV zIcNieN}Y{#kzxl+bi;WaXFKH5L#z#o42VKGE~0-jH46HnV=1a8$!R0$(wXQ4M>hmL zov)_hdy4^bt-=mOxt$`PB0OIaIYk3!v9=|2nio7YDe${!_{Tz3U+RA`Sc1mM(tnYUDH;4=Sz>1BSYjB0^(%!KFEp(U}_MwSu-3UA~FS99Q+*{=X_Jr7fbf{St6 z90|*@n6kr`T(rtPtE_<3H37o$0ETmxPEL(oK;3{^!cbTY+AiQrnS>c!&~+zP%p(gj zdZR1vT!0FJ(dB9lzTM?qmh6QL>Q-sX26hf0a#%LFR>U=nasz4bX|-p^=muY^5T-+s zNTI<%20wKK$&Op)FSY{Odawe%=3B45R={00JCq&ZeQ3nk`yA4G@TCEKLf>I*0fige z0@Gw3_9c;)q8Q+1;II0(?1UDtQ$U;35zGV&_ej9?O^szKEJQLZEs7!n7>1^&NeK+7 zBP>cnfeC~Ef;}Uwsue3F6urb1GPI&0Q62vA-`BJ+jjF3+-)+{iB)>AimCTiIbtnz+>GH+~j3QC- zrlu<2l=Y{8ayoXg9Q6X62$UJMrNoZ<=uMYWWm3!JH{Ocnw64W9Rzcf!{kmkNum zDTu_aHohB*aADjG^S))`&^RVQ)Bv)*tmCfN9vzqAF66ADgaj|KdkEH2jc%!uc_tG4 z!Qg_RumQG4Mxyil(!~!foL){YUA(xQoIksSv=YNh9F0&LcFvrA_JyT|Eszs zdVYr99(LyEmgnT}^RVjp>EtlV7Of$;0hjb@iAt)TOJuiT7}sL{&SV}ywA#1^H&E;E z?Eq=SEC{e<0izZ(zqNZ?A|3Fq`o(cF*)_-<>dhsI0LqJ)%xB~qs42acpNWI06&8)y?LCPpd-eQ!JxSuF+t*ENTF5Tm;saM-(&LMJBi2skp zO2ifX1z%?3hYlTe7j>QKEH2B!-^w9Bh!aUUhbR{xV8Ufwi^UelidPVh-M-P-j6FuNDwZCTC(+|}8(8}eLn6pBUVY!?!SLPu(Sa!TmnghR}rYJ!gB zTn%?&DFu#5VU|FdXp+|$a*rY(=pqP}{OsE9RRC@gYZn7R%<~WFY$;9|K|zs7k4Ii~ zd2^W6Bj5Zd%38Cu$y7Wl4Ge(uo0cMvt&G1qu5+}_LOyBQrHgB5i^U{y-j~%V7WXC{ zl0D$1qqJofJ9m{`hBwoUqsphF5>HV6omCOajC@3Mv5OX3%s z2N*PwfeG20aORRB#?gmr0=zEOSVCkAZVIeBw#I9 z#z=Eb<Lihz%;jQIXn$dff zDAz!z=+r$+)MwHfYx5+hnI>Cy&l1%J^PVNTXNku4EYUqnbk7o12KJsMs!ZiQOH^&m zx?7^!70c0e*Sq*(+a~X;ksp_9D5`TVZ=6!c1oo?cr4xWvBpkmpxZmV`#9KJW>TH*`Yzz$ez&1wNr%3={BdPFOB*lgVvXz_D)kdVT2vkoHyZ{@Md*eTuYKVle|qOsk0E~w1ugv#l4BuR4t#LP|hfm7B2^06wY4Uw=ZRJ znce2W7b)X@ZTI_kR8%7tZKcbaw!FRkwGWN7j%^ z7;f#X_2ao)ns?U_zliZkZjozZ7f%jT5zk5z-Y|CM0pyT)qWGDlu*_t-hynsY89mAu z^n<%_5gopLK zz@53VyIA^2WI>T45EPT)4)Zu24Uk0Oh9GKXme-wPN_ zIHv2n&3kajhkM}`lmiT9xA`vLFFp{3-@t5d9%d(2z}pXbfAGEByq89#1`g%Gn<6r- z>^AQ%!ui=CCq3tGZ0^sNcAIZQI4R<3lbLLVuCgPg5wrLpuI3Wb>XR5)shFdJrP9yj{(N++)qVKn&%6=3rU?y<0ZJ zxd2XKcCYR>4_C64B&Q(+!A$R6^aQh52zAp{2|hMAHxrd@q?0KnC~r+BdwmiuYm>Xp z`{@mWNYen$KxQ0~ZXQzIa5|bqK3iYoZgTaFBDy$) z>;aJdOg*=I?vCcWAW$1gJ*hPMB7x5CN1G2YHV9*)P^xkXGDAGd?0%$qC$5C*ub6T7 z*0xLa)K-!rpoY&tMrU%+jBk=kaHf{tPYQ&pYYQFC%S8;h)l~`}PUi7n$xwrg@R+U1VApna)L~agpg;WZD*)u0^J4 zk?C1vTJRSnEHVv?jQ=9zzQ}klGR}*P?;_*6$apR?j*E@l!-Un31J&GJJQ}%kVkmx* zMuvyt*J)&UD7I*1cql$iBf~@S85$WLir=G=;h}hfMuvytAdL(UMTS1{P&8>|cqpEw zk>R1J)5!2poTHK9p*Tk)!$a}DN&1xr;0_ui9+cmE4xe~XK1*Z7gL0ml@SvQaG2%g? z0uy3T$j0J9DR3-7^KM&Xun^y!Mp6OR5D;wfgtWS9CuWQNgmf4`h2Gd$Ef=*C;)ied zJ=CpniY5#IZgxoZ%HNK7;A3U3*2xsO-w#Xvvaji2%}JveF*MF@^G@84FN$jrFgR{M zEuVisWUuhAG05j`seO2NiS&F5w=R-6f~#TK6CLX%$n*G4{g^cf4z#f<$_0dswC25Q PC|P7_5aL^7HSzxgMx=(X diff --git a/doc/LectureNotes/_build/.doctrees/environment.pickle b/doc/LectureNotes/_build/.doctrees/environment.pickle index ff0a9aa96256665383b85a0284e3ed2b6d7f3c70..d7e6f3bc0f57106c18b0a02881b1a0d5110b8f06 100644 GIT binary patch literal 184642 zcmeFa37lLg8$of9=3xOnT2_z&WA^EcL;6Vr>%gf9ApE`G`d+YYh ztaYc01qI5J9K8XWM*HM^bp%|>2O-DgY{ zMDn%jSZTc5S>}o7|2_X87XrFExO9BI`?fKIbp+}v5-aO<^Zsa0!C@kbuDH^olj8vTq zXx?t;zEibHx8WA7(JAX*$0|15vHLdr{Cce*8XWDi`2`PgvsecJ!M%gd3%WkJu^O7a zLG)J^YZLBB9m+9Ms-ju)-Df-N+D&&vW>x@(rd&$}=pHS8fq}&HRrI{nY_{E|hi>LdAPlfKw&*Xs?p*&LbVUfAqD4(TSQnk{s@38w_zbbOUAwMfyo z4b|(F5}NLi({ijKYUd`@)2S9LuSZ*cmj*6QyH%{A!PlOsHCk@fda#V2Mf)LP3%aCK z^{GZ_ya;WdyL0=_9e5;Us?^47-OfV#F;K5UH&?nB^Obgi+PH-_2ValF;;h}aZJ<=2 zs*aB2i*EjWy$1D1<189DbY|oXR1y3xI$LZ{j5Y@fqhNw*2R*fxAMt5V zpz&df&@f8{7h}$a60CzzAn{<-X}Z*d(7-F8)@%Vh`#~0PadE0%Lytj=fXHG1?vjSW zfB?w#1n)$<(n2T6j{u)hw?SWfwzHlZe8jCfqflMqtlAW2fka$_epjoGjJY5K>r$vy zDh|}31+8B|n3u+;fLOCV+63J;nG=Y&QV;~EBTY2S1inwC&`U~{7TR?#sXF?$(0Ei| z?_PBC;@uIaQmIXj!0go;LP5e3(rb634!KsMy=(1Oz1@QKgIxi>bDPE5B>HApy;YiU zQK>t8b7z^?qek*D`7SlyCjSZ7cSn5t1@PuVXTam8kdz0HgnVad#et1PlZ}_EXqJWb z!kBo{$rVqBrY17^6h!u!)#K3F{Oi7Xu~13w+W(*ZFg@ZZ6p zhi*hhp<7(v+2|2a+srNAUEEt-eX_WwxVE^ixW2fdxUsmixU0CkxVgCJ+2UQ7i}w`w z7w;|JSA0S7{^A40q2ht!!Qz9Li^ovmA^bmFJYIbG+2X0<3yY5yA1l7(a`D{d;^W09 ziX+1}7hj4BPI0uDFBXb!XTe36kt&WqTPzk!opo}njWpc~d;>BipBUhRerv!fOw_6+ z()ocxqgF4}CaZYba>hrB7`0)Ep`uoybYTGIvBG~!<@OBao{onESm$L5{A8LMBlKnQ zdHJ>AA$p8woeMDF6D24p*>Qcc4KZT!}%b)o2UvufX2V zoPmh~Y#enAn>xmPw)EZN+uNOGJi3jH)yNbI?`%PnzXKb2hJUkehd}c&2*RTTp(Ca7 z0r6;Fi|i~K^(_4qj~3@^1=k*})bePT#mCVGFw9iamdpm68`y6LhI2zgpBr%MC4A)! zDxE-`z=@p49ukcX^UxC;wQlj7m;>aN504awo0Ym~|$2 z&6`AMZ|nyhnQMcD5a9)F#0AIC(=*SF2UC&cHc>yJ9^mERXNrG-u@cE{U?l z_1)5Aw__i2^Cj^m0=(I89me!%f)w(E3!kaL?S-BybaBRS?v;q+n=21i+T;-f0;$05 zY&?WvS$G$4BwJGuxPVSaP6O*iZz%Rjbnt-{K0@22=Wb`+A=j-34xleegt+@mceJxG zP~+kUpnqkWa-t82qCla!Xl_^}sWz7_49isAd{2_&UZF`qT=(wsD1iwSuoPTw!}yLDe0uvOBEe zH+@SVoXE9Il z$eMPG&vllQ3tV$$C6*Xb-N zl$!O5GZiU*2^BvsitCLUoc88vcZ{6iHQt|6(1RTMW@iPDpI}KCbKzx!x5b!67vYnD zpKft~kGW6ratl+d`4^lDY$K0g_)Hn527}fqL7>~2gV0O&Ynspg}3Mb3g4lYcQ7GA z#6@@j3;1u$2!kpcd31%3&5;oSU)7<|$NHxT-b({d`DK@$!nky)>%Ut*3~`Fj`F|FF zxcH;~@9=WPloswi#$}o!br!W-WA`tLxXBd3+|YMSr^m| zqh=`HMw$?hA(TSq5B_ii;T+7|0LHVyyDO>AM2*H<#13d46_(x9YLq6Rsk5r(6{TTry@JO-LDY7w?7rYu2& ztOs-=bVR;1rV1m7bx?fd@oqq8G2G7)7k+kaT+$8k6x}iUbhSVny&Wyw8o`8fT-Jx> zb?AeG=obf0U`*;$A#zKr2#6wD$Wg+RC#mbMxOg2?v!Dgmc#blhWBfK!aO=>E;cNZ!;*S^K;s4%Q zK_=CUOHdpjRMpNZQRKc1&F35EkgFyBdm`QxkXzhB&%bLed6+VErT4rUUl5Kf-XFZRLVOAn>T0VrR-!2Df_C+M6_f3dGPA?Mrf&KFlSN!8 zUT-4~%g?=Gti!V9DaPL28niM$9Y=Lndg?LWaD(}1@B(v(&@+0ZH&^w`ik;%?0x7tq z{YYQDs@HCq1bba%j!&e+qEQKAJfR|UL({AoaU5vgWs=#KjhTX}8G%VaKbE@ETG1R6 z@LWXE$YiNNT{3)^Wffxv0(dlqWmhI_Y_1`K#S|9@Q$G4D1#v7`%|SbPdE`L9!aDfs>M#Ww`Mzp?lx|9AfrG&Rlp zeD!r6F(0=4@k$Ej4I%(a^t+uiS4=ky2ZmISnN#%wiXp9&=CpjtDFzu5eBl)n^rM&| z&9iZuXLXkgNwEfjw^;M=(2+a}p?4T@u^8m}NP)tv6J=O4YmKz)@K9X1hr&1ZoivVj zHDQ{th$Q??zl2|fvnYa*SXB_CH+iIf4PM1z*8q{oLWvdyp7o0!>l1BL2Hs-shqz< z_S28@1zptZaztD$W4gku5tF>=6EuL|TPYIh`AA_IFUFA70&rrgv?>xCh8u}V)3SpV zxJDGWp^sn!5_Tx~eTb#P($M=@ez-v`DmCH3kGmIXedAWCxK^!9!FxojdNgQO3FJDP zq)H8U0>hYBg9_J4g>2?|y#b3x^(mW!{sYYm;|SkNHNzP9DhgT$zP<&wH7XCsJ;m&$ z`=YPTyXG2y9%I4uJ9QUvtiu=UeF&+-j&HS|T5RFJ239E=Rf``EU}3TXH>_{)B?H^; zy#I;Yp1Rbx_4{9P?(uWy9v>XK2O#X)m%LmmX9XF;=%-abfrR&DhQ_L=)rLVc)4KLkPJLQkywtTCt3bo(t*5u(7g zj+1_%5YcRuv1V|I|CI0$fLuT%l$CCO@V98#{ag<5^9>h<$g4%V4Rk3(0xCfo?KJ`| z)B0&jahv@qdxqs0v_vC1a3Vg+44l+7VFMLY4Lc=*1mylTeoZ#WG`7*Gjct3 znt@(K5kn;pn{ul{oux)UZ?-(3{KkCA<6#T2c) zDYg2xoU3lxDmRH}a1f$b2E-qIj0uZOQc`P*68SwUh_XV6i)NxC`j6g=C_wyus>B*8 z1Rbcin??T8OLjgUexIr-O-~Q$KjnQAB=&B=RU#E)^VS3N@C+o{H&mSp;~SZAoxO?F4K@o&&4V0Vr!yX|la^G3^MfT8%GxmuS$44GKd*XPlf#1T$ z6`>c3W+yaU5b5@&ffS+$W7&bU0SPa(EkYUBo%4g*e!cP$nx`J6e|%(CjJwEFx!9s0 z8R-Z7mj&J<5xrRHms9+g*NVzNcpZlumilFFn&4j!p0DBxRm_={!s|lYRy|dax!`2Xh*mUHP%zCa;4~JNCXkABeCXuSM~05#)#Jw>sFW!8Eqtx$7AHw5%a2kMJ%)dTiZ13zzer2+ zf_MA`DUZ-C;mvdOt#bnxkO$4biIg8?@w6wZUfj-GH=rsP;m0FBCO)!>>$ht4ihIGu z^qAMNVP+85qu|b3t`MI^h%%3+aIiUUa9y>=;lwrknkb@)38Hw2NQk$PObF;luv{Q= znc!R_mj*ruEkt%-^j3&3%n(`1+i95>tD$hR{2CSl)aZS}vtqB^ma<4}LEE}LpkSSY z#9ny$CC8yk)&8Sc`#k^nQ*Ep!HblCj+q@Lim#RP2t|Ai)YxLC5!Wy?*h_R5x7BK`d z;$uibP$r~-Nm;9Qj@%}M`4PJtr`Xq1=U@@sNt%BmaqX#d=%wWIH7T-p=_$`+XkH3G zbPkDKh!0Yp&r<{#HRDgptp^*_JRx$}L^Crczavpn-CD^4l>^ zo5IHpmofobBSqJNJ*FwNptwq|fjJ8%Si{%GOZMb&UH`R+I7VHi=YzT{GIYIp6kr#1rmfkd0EbaV8#2K;cKygC*D{_Z-gC> zTl{)@IN_YfG?}|Gt)q!V0zZ1chH?!nF3-*VphCGHmNhWVtI#=1)iIG}8(CuXc@jZk zgmJ%!w>q~x44(x;y6Bfd1dvtCTUcCnqSU11BSc4P?Pk{(V731l)tasJizE{vgkJ9# zBwC|@@vvDs3;ne+27R4RpbGJ@{#&ZR1hqZ8MDHWJkXw$71LB|yw%up{z~(z&8V(T4 zL@m!-qwFPEiHl+J8c|G3c!6(2ZjT?NB?iPdI2AMM_V|?+Uqs{t_O8{ES20%Pkyku| zOjHp@g}O?k)@q72!o2h-UY_@IEJ1YS!Y(N?yfH7Lbs!$)n`fSyq z`6;s4^XD;IH;a;h>!4oHcehf9M^G(zYpPwL!<|Jzp>R6Z1b_N<;a{9Aubsnw$J(&C z;9j864XisdTHUG*y$-5v46Aux^p|Nm*X{_bEf1@SeBZ?HoFAnC52eclmcR~4tPDw@ zBIO7t5~1lJ7K@|h8^U*awbHj+mjgjIg;gd*B5q@1q=LZ=RyO>O)MiaM5wME|w8fXm zk@4G2Vv|?%W92Kt?}#B48HAqEdl0aN;@%L}qhXnFM~|QtBJ`7xBdPs^YP@cW4&(E* zKBOhY#|*@amJPO>jX~^C7%Y~Mos8uN{}rjl0=oRps^A^JAbiUTKLjT^NVfLUz-w4* z2^)MR)QUAVvfxK+^5VHr(p6-;qqRDVc(f2{Dx%wniIb?595fBfX?*gnP&lz#@iL6e z)MX-!1VhA1uUMofz`2qa$(MR%VrWP2sMfFnBAj)sm5URm9>t5rDgQg=V>B*3Bn?jG zVf+!!%gV>`XE2Q_pTJ*11h;%L^4G)ot9*+7@unc<7t$Z6r6-ge>>FxI5DtDv!f^0A zl8A%fUlL{joC|)H^U>L4jrs;_73~>7jG;y;9F#;V=Z_%2bkv(Oj+}RqUq7EGkm1m> zI;?k=fmWufsVRE*&vq55XwIJXn5AnVSsdSYjhEmAib|$dE4=H;TV_rf*4z-wD zP`F&BiE*`@g_pKms8xDx)+oL(z=3)6ZA^0tYo{XVBv21-KWTeL@v-1#k41RMa1XJO ziCx$*aIsvxEJX0q=7Lu@a->Ajn9kki)f4bvxHy z>bm~NkftB}`f%`TXDMvWs0-)B%gXCRs%Mb8KYu`!BEQ^!#!2}|bc3fuyBoZ!Ug|vt z+5wJjg&{B6Q{dYT+9wb&@nF~w)<|0(C}X=$Q)sDsZjj5xrW8bxkbRFSgy-tW6@rO} zD8sG?|C5~!(XCbRfxMknAzEv_DwH^ncz=mTok)YP2sQv*;05=9eDTTRGr^~Bls+{^ zi}90gleV0Rxtr8?Hu*{Ym6}7rv~eH-pNC>x0Z7Hy6ki(<4qou5$k+)${vkSiAVnMb zm&+JZvEU6CYxr7Px+0LhC$w(~8Qrw?48aT9hIO*D01gvpG%n}xz(DX|!6~mONR;E?{RNMo><>1ifkK$A&U)coV!tN(I`qLCwh9n+jhqVSr|>@RnDFc6rY({PZV+XS?~?9V^yi8$nu)@a!(Gx8<9k{Alp( zZhqGHYft}O@N6$X+xojN{;}ZMK7O`s>sN0LyRrDhwr#_Q{r=1k`p@nUOawn^Z5)$K z6FCitMX?+=MCNRXZgqH_{Sgek3Zrc zpXVQ6;2(d?KmLS&e35_rDgXF0{_!RL@#p;GFZjn_@{cczAJC9eRczr!BH&GP;qP1c zhdAmi!^#ZY9k{HM&Q!De2Cz<^4J8SCRc}Vs&RkD0mSQm;QMX@=*dOKk)Lmp@9e!%F zxBMZ}wKv~mZT6{DAqyaHWAi`c< zAyRb|Hj@T;jK8#5$!$=LkeDzFV1zS_sSNfFBDTv*L<2q#L~WLrK7T}A_WTib`SVBA z70(}0S3ZA4UG@ACb@lT{)HTl^QP)3zMBVuO5tZf{&kG;8+3w62lWGw_DASBWOuuQZ zsX0I?$(TCL6E*li2AXdyYB-Z#f<0ngn1JGpGKBhR1-L{6=AMPOP1CP6VGJoaCP|#9 z04n7wRZXmcbOTyFZgtBQ)M=06_$6tbnK+nJ!m=JF@O-`(c*O7XI8=t;7Gj%H_y;Y+ zigX{8EDBb^@y75Hr;Q!c4%TaFx2`8;dsIvd%IEQEKGaQcinW*;ozr;UY z&Oe^!AFtpauM$6ompSe7Gx)XqEdGH!?={D3<5-O;<$E%m zSc5tJkTIt?d@Rz!DC@cejv#ob8OlQBecFW!Dp)kJz3*Dmg-~1QZ0x{MgH;( zQdz-KBb2g%qei9|X6EWF!{>};SQwLm9N`&c$R9>%Q$8N$9!s3q;#g+@NyWA36ctL? zKH*fS#2>@>bq*@YTQye`6)qA*E zsr6eY$$OZJiktPvxYcgH!d9dC&kXvv|2Xn7!xOLitp|~Cj0MGUl7FbtK)_oJ=o8ow zVVx2&VE9xbhc2e8S44^B?bnvK?0Hq!#q@f^%#!!{TX!vgm6kXCP?i(664q!BAx?+D zu5voKEEr1Dzs54{*On=&`82Vc0mFL_d+G6n8P=^wol)!)f(oB4x;XAu;D_9JJ~<6G zG;%U(RJto8#S$OWme@z}+8BzPW|n=eKoRNHLdv5Jk(CmaDY**s`XX&!qv%{eGlz?4 z!o^GY5k-=oGcRLT$P3MjXV8>YUT^Q^m?xAoZ>=bAvoyz>|ESzZaUdzJqBp95JqbkW@(2SjD z-AhK>m5R$-+k&}zZ0;(i=Zi#M81{V8X^Kfx+&*L>i9>pDx@f_R0j%(%4vG_s)H^k+bAUW<9J ze-)Hw9e*0Ct@Mb`T%Ec78*RB$cFWGKIUQ85Q&IgZZPpZ2|H72SpX0?;aQ2z@=e3XV z8AaB^?S}R6WEJyPq$bJ6vy)tf*;+C`tG>A|=bfHp;|N0=XJpx&G{t^2^}snQRtvS+ zQLvgPNc6O*v)-YWvO(n;OSIvmMcm-FwuEUr$}-Owa>`3rGIev3jNLYEMwF)BVoDs=fG}HfM@zz1WmQ97_XrIE`;}6SH?_KqBPg4W>f} zapA8*^>Pn;8Z+bJB<NWZ_iE_A*>CCV57ChU6#s&MgF}aK%8P#(Q*;!Wt|F`HhqdKT+3KAea^heEa4CFi z;?fSd3`Uq`N#n~eYcr*Q@`rUgh?6t%Rwm7MaE{+;2@o^t?L=SeY z#3n^)Ut+BOM7vqiVs%H(W$tZe9y1`BdD~}RhAj!W&ueq20QWgl5~1DvG-E&8)8Dl; z{W((`YvJ6{9{nx33ZwU*+JdD;Z=IOtG?}yHJ~o^%x%pq(3@Kp!t0{>%SO(){IMi}q zHy{ym@&4&m#&}E|Iz)%i)M)P<95|ZA;C!xzRlqq~bV8QM^zc{JII=I!{zHGm88f^o zNgggcvp|ovar=VVnY<~ZZRY^&zArPJK$6*p%(jr(W|{FrWVbYJy6VIxnaRckGD~(Q zL2+$u-WA2&CP?(0>GdA7O^1gY58}=-Y(Ww$0NAHV;?nI1me`;X!Fc+r4Bu((CP-^c zZpvZfE^%sOjSXA|oZ7sV#GivN8@?`S1bA4RF$J6BrX<4de2n3p?Ga$ZSsZSMLpgEg z5S?Bxu*0UH*cxNcS`6K!ws>iw+nB@V(Na|aNBjxI1b=dWN*W4U+B_+!z08zE=BV8y zyk#7>3QJTG;E5ch7Id60wTkq;8R{~EKdvomS_E%qpBX1~5w8Ns2(J+w6JlU*)82@N7_O7FVaD>0+6*ax{ZCU8nFF@$(NeRGtCW+RqF5QjbN!;M<{`^-7b3h!s3c4{ zHfZyqpt9DKMCPd6(9D<4ms&P1!Nr9~7)dyPKf!UGwk9KXzqXiZJJgLinNK^$r-Wa^ zT;pDCt`z+CnUcsHzgW+S__47v#;BkzN?MHO=1K*e6df)dBup+G?XN3XJYh;Ab1b$8 zo4w;}eppU8i({o~jj>8?QbcKnu}3x*R?~l-5xuP4AZZcZkrTJQ(iA_wqK5?tIw*@J z5)m96E(VDt_!DL-uhwQ&f&UezBr=Ep4$2%jfaBS)W^|CI2Mw{ksUJ5vP4cR50_bU9 zVyK_f7CtT1cZlqQQNg{P7u<`5B^%81@~wjJVs$3^UTyvqeBWbAB6EDV9l`!YyM>Dr zlvRBz?&n49u+EE&^JldiAuY~#=I9z4Pl{Ntp3YlcqJsS?Z6+14f6tUe=3sAmtXRqy zExI!Mf^dd{Tp<6R#bb2m+WoO9WZJhF=6}%^KP}8#a+Acr9r7P!4B0r|9rr}qC0zf2 zBTB>JdZ`NWx3n2lApS>F5}8B1k?i&9!^a;9p&fH3N;oDy(tCSXV-RmzoYgdCld#DR zx6%&5^Ei}XTc*Nxy*5h&CC{!JrHp=eCDi!JHw0TsJey=Hs%#q%6obJsKzN9!zsM$taAr6T{;Y9F_W)+kt ziyP_OII?%PTUE?ve7o{P&N)MQUpQd0F!J%TCl7Qt5t3I ztTw-jHhjvIL|hvT)+EDaP2M_WKqBPgjixF1#D#yfU$4DOv%hsI=@dYD;3@xPupw;a zlG$;e>F~nD4Poy}+7Px(`>Tp>EfyrMVsiP?V{|nN-Ks~on8?l9i578pAJi5xxw|V3 z+gX+qm$dkt1f%!dq*U>KZ8nsa-fBuB>_2xfShJlfz8)8%(W8 zv)aO?g=qsaf0#QOiKUx&B)}5(UA$16AqB9LrX*4WX1rBLwoF(tF4k~BY@{nCRbZs1 zw53XmRBYvBZqmUA7qq{wz|s^XdRnZW!+>eh%p#m5*%CK>MVnZ9&Mo_6+Onl>*~K{= zcjfUQseR&k2XEEpLTTY2F(nZy{tugy2>lJl`HK9FjO^vDk0L7)*C|^#c&FZ^V;u91 zTC-^*-KB($N6NUCqCRWTC$(EGtp;t*(Oq=l85R^g+gSKgc$PFPAr0r_+Po=x_1mT- zLcPie(d#LQI$ClFL!3~ERWA$MlL~HaY%JR8r>n>qenVTxv>4vNi(4|6cEF_sUP4OF z*R+{Z!260Ri8y!$JvuZB?5-M+2)S5bDu_1;5r^bmy-CP)M|;ade)+jzGIHnA*_n)x zq=}P}Urw5g^l5)p(XrbFiJn9I?C*3H$9^?2ucy>=4Nr~!J~)z z>7D`ycczLn=!b%1@giW(RGFcyX$zYc%A5HlgbM|1JRkF!O*+GGUeYeuiZ)*gjwMqP znd7+fFz&dkO-$KM>;SXQJbEZP2&=`6RaaY}v{2%5!{Lu8J56~NI;?am;_R1Oz?ic zsLi4R^A}7>WDfIgKb}h1wlIx3SPSMbboj%;S_8Qk(d3_4u*i%A!#`(PM!sfePY|D_ zbr(p6{kFbjB2_aVN)}|6NU59ih7+>juhC{#QHB5bj#Y(~IE?%}sexF4Q-@8Zm?7-b zZj-b{qT`B#Gp>ZmqG$r?u@CQeZQ-=qoZi1t?%Ea1 zHghNDT+L6KY5sur*Olge%9KRdi!L%^vpv(CFHS>_jidfUZvA&@GoQBg7w`$l(pWb? zX^QwW+6*WFywj9K<^ZgtseZ-3W-d6ZQ#LrA9(5R=PiTvl7M}HDV(;@7xF+F@aemT_ z@prWOQPBFBDT&O{I^`WWDpH0jrSop3gddaZk`)I@v+VNrR@QZ#7PqYeYTZu#uYVibp@vpzTJ^CoXdl8r&a zjCHj(KMGnaOi5&p){u8PkeHM9^RduKq^0hS9q=s*P|%f*Cmgzb)3uJ%o?e>Oi0e&o zmr?w-mRX?3T5ykclcnwML;g8KLS~-KvdB0DtkEJH0Fmtp7CFo|ltd0f9P^XrwRdZ? zt|-iIQxcgg%$CzGGEA`^L(wZ-|8d+O_D=^E(aJl`*g+<#-(v7bw8c*gejhTJeE!8H zAxRcp@GoJUFgZS_&7p$vi%m&nj`4c(=iHa!xD%vLh}#02waSGUq`fLJg0IvTE-iu^ zghv-}hdgncxsw1(*e3q8HbV+vKVV8CbHKKo0Y20t+q4>qH2=vK^7BsPsi`c3h z5_|}=@` zTf&C#S8B7Q!1lB$iOgYJME(d}b{o4YUV6^JyiZ$%w7@JDehBj*b)bZe-9M+zgM!4n zOi5&p#8yJ0i6gQQeYxOP=@J-(=URB^?|(xj)@({`kpDl>8qXbYRRTirs)dEC;DTIM#nldV;Oyic1k1;~3$Nn{RW;0I0* zVee`To~tcjTJS=5aXQ%6sj!V|v!uW_VoD-&*w&swfW+eHgMW8vZ1_s9!f?GtTd=fn ztwWf3!o@BII*^!d(W6^cBWXM9Rj8iPW=DbQm8K*zhic7{2DZpfS!djO%bgfSa{A6a z(V;A{1_SkeZL!hSv) zoU|yd3=c|E*VHzt*!;6LCki&-G$oNaHXGy7LfqgZPVb1Cz~0put(#Y6<<@UhqK649 zq1UUZ-Js2pg4#w?65lPT9n=;uZ6Bj-?49yxb%Tn~1KM9#5W3HlMCJ&s!kN~{9&;~Z zZ4CF=BY`B^p%TiuCl$3tN{b1udL*7~nJ3AgxKTyO)#gP(Xw;NM<_M`K&AsT5R~!YZ zcQwZAb=m@^#cPvl`Fsu~*lto`dyO_r3T)4qlE@skO$Te$avPVsOYz>M5o?-y4BQ8` zMNA9a^%T!w9>aL9>T2AqBKLl6o)qNXXG-F`2f5E{iTINcjmNMf+6|Dlmu>AmhKa2{k_3`7vRQ@spf-yN%nz87 z$Q{`0l~zm$gOI!Dz3H(d{ZmAJqQ3g3NQG+Aq$ zkB(zA)McCZC)%Q>#p>qx_9nrr9q=lRSa+x>eqNg|1;x*qlE@szjX^BlOOir}DkA{ z)@%({aL{kCZ9Uo#)4#^B-nJ&I8NpVO@)kdJ6g`L5ID+#TwYNtG+&5Q!lK~oZ$gS6b+4TCHSq-{<_0g4wgI8|+7(t@+X-|OOY zBLTBb1?IdqBMLA@Qxcg26Ap5RI1|`9JpCXSuf|}#SzExgV9}&(pX^NDsUq}7?XN2c zz0QM*%kLl>UbuyWT+W?ca^|r!+4xY#$acn87mLEKh>&)5>_hhZ*HV5-X2f0g)-2--0nj=Pu<{+m0puj%NIAxLKQd5ZKiWLTv}Bh%poHN=LtuU_FXz}uX7X* z{}kK#!~JTn_+`TeFMlX-96o4DBEw$6%6Glub))U9)1|_AY*BK$iVWztw1rFy=nbla zP6#7e4O~itj;(_CkJ?NrChdQklE@sqdrq+%5LmIXT0_~A!slQLM0*}ld?kb!{Oi|d z<;L&NdG1BfgCS+O$m!46#FcbHBcvmuWaSu8m1Bc8(~5GeH6@X`a@_fdIFbTxqPU^c z-=XB?S~dGE>7KvnxTp0d!+*baGo){Rs0-nt`z*9 zFeQ;Wej5+szKu4YX+W_SU*BU%T)nF?a+kFQOpDwmEYAfTQY9JVEAjvn%d-)J|BlrT8YX^W8-oTa(uc)=easAQ5u$V$&0(y(=wNAf(#!@`auk{~F)2y!A}GUa#Rq;ztqxg|d2~ zVFVCr7m!lep!11wEO7~2dgU;0`IE$^ST_ABZRC>Gk_1n@=;E^|@Q?}@YkxxftBT6~ zm>|)!t$W|ZXx!KvADpV=aJ!0wLuN$UrcjtOyMxS^m-#rCR&p1cIXjfXLdZmPA(#Y`E(Rq<|Yp*sGY0bc$&@CZFQ>nIsp4QJN-Mz0u&t>mQq$C9#C}Ly#I%TApF7Vy z^8GUr9wp=wE?RxLHctw2-)~AHbL9FC1#t+keT4sd!bPS@L!5*UR|O~j#&Xu~nC>+k zk?z05sJ>fU{IsZU$%XL=8Q%%PKhDAr81ZvO77Qm`S@W~n3@Q-+lqreKAr`Uoi5lHZ z;RQ*foid@EVf&=ENNHi?*t+0JUHw>)bVtm`wRuqx`fXDZnImMMaGjQ>oQk zSQ~#*?E0f6MLsPp$lH!NG@9=+vft2dinPcMQMKg|Sv!b;6AyX1z&l+OjW;>d(@+i)nriAiPZ2I2W@XS&@NR(HaYj6e(BoIXiu89 zCZpEXmNPACH{?b=KD|Dbm?*h0=~~t|X)~qZ_j*C%DjKzACmeANe3C_rm2t~|NL$FX zEx#PT4Dpg+%jP6mvtQEYL}~61n34#)%P%luvu(|8K9l#{Iw>95!g6TdEx@=%JIqX# z8M!ZN3!4_XTXIdGTdB9uA+{d3lHHq=Wcj|J&6)z`ADWVgLuueOhK?uqUjq^$7c&fx z#`bsJkaf_D=;e8`bCx`YNW{NK9oCX4|Qho}|@aOx!HOpg#so6q9tXFhm2|O_H{Y-kII;%jH?-hv5)4R$Tx4az`-UpxHAMM~6F61? zIVNY^mbmA*T@PpfWKUvRmi{$XwO4B6kXF^U=ENn{LIPW)NU;QVyN(Sj9KR5jKca65 z0pzE(SyZ&`2Ly?$X!>tI3v+*P==52>oSYg#pcoyLqqaZgJB-`Av>PQYZg+^>>^$8* z9snrtiw@ed`Jb0${(nZBLj|^Xnvw{+=Q|kr**5>T_P{s1csz=2iu#P}C$vRQi|giI zxHf{z(xaFrEQSA$Hg^i9A2TJ9Ii}t|BHvj%TB??Cv1-AMo%zwb8YB31Z2{9F=x;Zo zX){J~dmJ3%Dplf8nsd!didKA8n<)jje=sGHIdJ!R+b2#9n39O=iGf=ddUibG8;}UO$jY<3>2MXtVbrIDPNE-M#}4fZ-Gu14 zF+*Kewx3;4ur$~qzL>WF#bS44?{EowJlT43+Z<%OJ{!5}$p$Z3ls z1ESu!fSzZKe0qAsU!R}k`nlSCE5MJMk_h|n2{W@)d(_(Gm>=(w*-^WKCnG-~qVaa=#pETcpjW%xzp3j()h~sHs(!)`K&4~et zkc+HFg&Ps>=Nv!eaqwljG`sJod+TDBW%{bDPQRwz-f7k8M%wq(6jp&ar6`h!cF_tz z#55%2ocxM5SBfHiSdi#BBJ|wsW10xC>`0yVG(fi=7Ac{6-CW{ zyzjQeIQdXpyoj`#B&VtT4?(KZ9oNs!9EzGOF;(dsNvTRX?XN0b#u6lY8mQho95Hzg7F)}$9B{@F))|jt7CdEa%G&Iw#bzOr%=r7k*s>_eAspA{ zK*6A3N+LA|#*61j?h}pRpH~`lztm!eGeqyyD4o>7343NlE@tSeqf6IpRfPP%@;(7&GsRN9|1X%)@Z4-Cmf#e6WVMlkpGw|i8$m2<9+Bd@PK1L zBIII*`GIMKZuqEw;S>U$=YipC&ots`W1m~Ii1C~<>4%PF^WJ{KidhiF@LMTf8NmCM;tJBs% zCk_Mw_QQO&j#xPCm4vAA9;?E)wVNQVDr`+q1#fj)tSBbsBrZzY#r{v)EGj^M!<5AT z6`*6Tiidl2v`PZH$9oL)jW=dBzfA%<=Ad}E%RPFbnoP7wn?(ia^`;~;2YU01j*6~) z4?@nobf$FiD8gx8U}%T5g-+XtH|H*n3p z6`>in4$|?1kM07tJvhVIZ?2AsVRv#HU>Uj z=oa#NtO1FTi<_<_o`q%m)mL}feh*J&N(I;9L?Yyvo@$h;`BEJYCB;r_RXQAd2xoUr zHA_va@6^Getr7h!|7)-i|4wZblU0%$A#ox88r%uy=}vf5?K`x;s;J-F1&ON|fZBbG zl4T3+yuVa+uyz3lwNav_2q_$-YxXL<9oKhlV`~LwB-nQRx^|PL1gf(qR|+K)>rweY zGA|Jg_#>ty!mj-xR+8BcGOd=e#tBsF&9Mm;7{r43?=nEzMi!h zuSGY_j_<74$IixFi}@2>uAaxSKOTP)XbBOW`Pv*Qpv@5^uA=R{@hk?ZAoymDy0mzi zjK}m<8I(Jl`x=C!K3wK5`r-{fUS-n1O#uRLBHzg5vsLc%TY+JpJk72_y4Ss$O z2lXlSGMnDj7`#*30;UCTQ*M$uwEcVw>Q&KeZ5AVqB4O*}32mkn;2tt1kvVX~hf2;k z9pzqnQgk}2=}gp--x&_`-u35{>{%Oa(Y6@aBIKs~UL{?GCcMg+Lo=(CSv}g?&6rj_ zjw8W3Xa`|*xgDS(z9Z*dxK4ZSxExJDUFkHQ%i z4;BU_LM}3L!`5-si%$L|Q-m=_9o~D?VO@Kdb_1u?we>j;gMl}@$IzaTtn)M4{3tr} zPE!(@qjm78TSE?T1IuujN#V|qhU;DaNNe{v_WHcHT!JB{U2z)LqfVvm(s3Bn>6#dN z;>-B6LXkD$6WYy|Ruc~8#(lZ5g;$na%&kanZ+*m*+rB_@QyQQow`6S+@;QD-n|noX zK4wZHbG=zheVbGya)Ws!+?HI0k^Q>1U}=$Em!qC7vN7_d^QD$u!NKiSdCHZrboEtj zb`+@o!IVU5sLZEa2OQii0WX5w2WhWuB>L3Z(5n_Bw(jQard*@ok4AU!hcGGa1r~c` zWzZ5fwyoCYNC9nyAaNC^Tq}-M^R!I|JKMzY7==J7<)*(|ThO$fWhH`40Y5Y(Mw&cn zF11^m7p2j6n34#4$^hdw+jA)|YZwlBfqa)JIPpTp=A5=LX|eHBiisN)nuOWXi?vx% zpn24kMCQ;e<{LRwSmXkR`F%gq<7rI)qiTZPqPR*(;BH)~o2S(W2W{}M-%kC?E+{(fy%6eW3| zDTz2d1_*~r!V5nJBtkCUVj8q2F0kUa=IXBatvKnn=`{4Ji;M3X=c6&0@ob~PZ5GHzck5%TByyIg3LTq5@Bb%PLP=O9fda@tW_`6Ds2i2pl^6x zo~=&J-M393Yr? z>&R!RM!ktnuSyKrPijk-7P1XFab>6A&X&285KD-B{0rCs1@U)u3`_eG9dS4>HSUG2k+{cJ~f)(!;?(d+WL{AjhK zUB%YJMiWyJaohRD66_?VCdo-oS4N7F-KX7xX|2S)xiOprWfFfHcA)#L}gm^qJt*uI1g0l zB&9@&3kw|C;)Gt8HGvA1r~@)PBw2OK+Kr!9br0tXp5#JT#d4n{xkv^R*Bv-~MSR@f z#KEZ~xunAh%YozC>?;~v5F~oquGv36;bx%`fd2w!Y_ytAHVcOkA=RtVuEe&l%v70C z>}WSbS`=?#6q#G$S%O!Ez1oySXk%W%AkVgUx#s}#(ZqpS zC7ilg8^`g~Zlm8CD>bmkSn2C7PKr=dXAEgD@V}_tG--j~lNgZaGq-0B96=Wa$R26N$;pSeqgKBW-chLTu&6 zh~F5Rxbq_5H-cuua_48Yc~j8*lqreK(cCKC1VtwLKz!^Da?_*3v&MUj@V{#}L0W{j z<(lDLP-M83jvLAl!8#!#_b=L9Dp-HZltdhB15Yn>7J0xoAQ5uGGfx8&As4*RWrzXn#x5?fs@CLOl1HlJM|cXN0G5nB<9}dc9IY)|9ySC)ndk9_*N#1V26y8h%i&|HAK|%rsWn9>jXvD0)cUQHv=}@U)r;E$ zy{KrnQnIzA)|yE#O4{F2^kPhqxQf$@rT*qv)zyp^@!<1XZ4py8XJ=WknN@J2bh>A? zzpOOwt4v9Ro$iMipV^)ntnyVKTeMn8ACIRgC6qHB@7ES7Egq|LA}bo*5u36+o)l2t zr~P>al%F#tkvS;1&%qd()Y|Dns>IZ9&sQbrY`skT@0d zZi!O`wNGhtqM-JBrX*6MW<1b{0DB%A37f7}#7PUtxUuN;E=t70>9p}4G@O@ib z;(f&8wJ8|2ohIusIkw`iX)E2qkUfvrjp{23vvKz zIR0;<^rQ%P(7;aDmV_H{1WS{oe6-4eb zB@su&V4fFF6}evwjqyBC^H?wAH7LM}Yh`4$8qItyC0JWk+qyKnAp zeY$(8+gGk4ics8KuHnby-SQNE(9@^*&rAH{<^1Dm{_zU&qcflGRj9VQW#Nk~Aa|Bj zJ zZ{q#V-0@0nw0pTTzgQ|1+-mpP&NAPL@9>{=JBXD0{8@r}$KsExI`H?zk1_W3x9x-#waeK@OZP!avis-htDu ze-T*V27AX0pLWG@7Vt`VSQOio#8U5ZM8&Y5eyB3;!Z^g*dj-Cz*WM@oc1}`*m;a66 z|G(AN3$Z#}M|5`wYp?g$&d6W8B7g0U{Iv&vB`Nl1LghC4=MMU(kN(+)f2IRzn@ALr z4v0j)(3y6*-!b8VY3Ft&s8_YfF&MVk2T9!-3Z+Vt=YZE8pEfF^^nxP5T9l`5ev1g@ z7?244mb_^|$y-fYPs8);;q@er%e{POPDOGQ5epxdL4j$5`+91MdF2uEjr1MYmI)Gbp}#FmQ0YI*q{ zd`fYi->Lmo#d&^*Aknk+XFn$`x-t|=Q;8PwB;~iXMNFQgC=DB5GzyXuu1m@-`*m$L zl$QReDT%P7eT2d4)gw&T%k5)0YpKz~eb`t|rF|Y^SoLuutw1;@2f$Q|B z!zeDjEhC%PS&z$6>Bbvx*U^}7d?mCJQp6T%^P`|O-;_k=XstZio~TcWj_HUYDl!sv z)MCbHhqgdzd)lfT=Hm4!v2NjUB*BxA)it2aiUQA_rX*6sW9(F{J?gY^5WiEk4p*=q zRZU30O|HT?J*q8PTAbG9#y!q#SK>^LNek81T%{d70#NK{@I;^mL2O`%9r1l#@^yc0Z6*}} z8>S?}&Q>!e5l&B7jjm{V!d{aBiI5BSk_||NZSi(fN5GpTZJ2viCrN#0DHKeTB0oaR z2X$hebRKyTv=@%T5SYElE`0ewFeCakZS0e~3CSIs5qYW_CRP55_O}#m|F9sTZ?237 zj1`Z$?)l1;b;b|C%k9`vO4;`OnYN%Q8@01CH!1kBeYn+C?myA~xYEj>pT3o6f3~@n zTG`sQ#~NynBXDNz+-2{)JEF3ksM`zb|`wZ5PTw3F?c12;Dp)TXK`S#hFk|;)}bF+unT|Qpm zwdwO}?^1iqE!wOoz};v{B6Hx@GjMlXAzXLaJGMs=>s5)NI-)IH+OD<1gX(SxAiF(xSA{Lus#s67#3PG_K8r0#m`1 z#MOdnp9Is+?e@;y(Z1xtq}F1XUau{lE=>C*n3z8Wrq^mSp}_R4DT&Nsx<&M+-P>iD z?zXLm^Q{^URy(3_%~+dZ`=GYCX<@sCd)IDxklJ~-jhqr<*WRIySMS&6MM3X=~i_%&Tr5yn<#Fx^sKC8`u0?nsPNn{SqI)-Lv0L_ja z_KrPKaC+2Xc>YaWw6yT7_u$zXz{9*L0DW7V1qGmgG9_`f0NND-wAbFbGZCP89R}#; zJF;@heSmg_014g{fNs!cK>=u^DT%8E(Cz@BoqOz^`w{_)*I|GTXp5%XgLVf1F>eY$ z_iMAD0CcY@iOd1w2*d6@*lv1(=e9fSn1xHO!tjh~3zilhF9@-F4^H%B&TQgLK_{=x zfr8FUO-W>qjyJg66{54(-jQtI;&m9G*Jz8D79Zcf-4)^^cvAp+MwQ`^{T`WJ*O>PT8O-0!|uC7h?qMCqxWj_pkVYKQxaDTqrD+UJ9pYU z?}`pW9!5$f#^^KJ!s%kPH^hj!Q!x4iZ5|YiK50rKbBw%6={^ageKxI`@|YAec{A2# z)c#pp+_b3qQ`mhHYWrw;RIqFBREMc=YV)F?_jOYear6w<3d7|_UM4Xh5pwZ4)1ZII z1uuzR5f{IFC0yX#n2mF<>Z0ephb!XpEOET~sYY$o87+zY=(BXpC1>9RJ20>jIdU9H z#MYHh%>(CR-^uZ`3wb@pbY{9Rrxzp5}B*Y^>ALTd>d!> zBPaQSOJ{=ner+Ds|nEi)yNIe4p2 z;RJWL(Si?5Da@x19f^(FNfj8Zy=jH|k=&=R5&z*>0q@9w#Jy>%H@pUihxzPLdUCxm zCf$&A?Qp^|>U+|16)>ZqwbPVD=4kB*H&S6x#!jkHaZ_D*Sp+U0oUr7gXb&`apRs;i zyGhbo-W}40Dw%PdG=>ox*Fum;wp=0)%L!MKexEk0iUK@lN+NRwSQqR9iL^+>&+bu& zf&3wD(b59BKHLWq;Y~8Gi(Id%!_4ZJYxAR^_5G$KGDmCU;frXX_+ zrG;t@LFIFWiLBsC>RU@xkUp->i~`bco07;Jq`P${C=)LB%_SjeOoie8hIaF$h5K%u z2?>dppb|?}mH3)A!-`6L#gs(mDsfClCHzxYCvc!~)sJ+D_z=PYQLkwh>9X3a-=2p>zmz$EvTo3lo3_EYfuOeD7?PTkUD6wArymm9C)rOsT7P8sLbD$bwO{8`aaACH@oc)sb!v)T=nRzJuO z-6i{>8&srUrTujU=^rvBkvY=qPPi`LHh@$)KV*7qG-}+V4)@((&=xH%HfwVe;7v>t zL)K)hHmX?tyf!xqR_`_?kvUddlCYwiTlu6rveDzXMlTlgPyH6-`5A5T)8e@$2~Qh& zi_AV5UuItl*l$uH{sV0W6^K7+N+NTJ7at5dVYCm*1q{@;wS`Cv)RJ6sf4g1<-eV|lyfdq*R9|ABlO?b{38YSU zgNpMeZ4MQj*PD{a9Oph97>olN>ApnDOlZ)B4=vtWLlY&bGenYI(!a(i4{0|*+77#w z4j5($Y)OJ3J6w`Lg84=j<`-zQsKC76l*IQ6%wyUOkQQcgR(Hz|&`m0WdF`(&2)@*m zMCJ(YPuo%ZEhpb;7U%t`yR*$=+LD-32ftZWk7u=+ zR@CEFrX(^~kA3OY12?N#wb7>AxWJpJB8L46=&)*hK)Z?3s>Z(bssRygSwu-OZc!EE z7qnSc6yxVjNo1}V_oY`18t}zEaV~HEQ&Om|kQOV*A8I#MS_Qc;y@F6=ltty;ge0n@ zB$lcqpV4MpQIbC}C6T$3+@D@a{ER!gBtQgDaQ)FlbPjOE)L2peO}n|$D$4!o6~)i4 z6XH@lMu;n^%B`xZd|R7wMOFUEltkvLa?c@L?O83+g`Kq3=%6V{ zPNRUMLoSpO?zy-ULaZRS^k+47+@GThOj%0WvE)yBA!V3{C#D#<##2a%6Gwb0CFR(x zD#wl5Oe@N<$&|$NNjVN_H&R;bK~s@E@@SD$Q6AF%x`OfxOi4VSP%ddVk`Bsu$td5Z zqCBSkbp_?TDT(J3%CFaMBpsCRmQlW4MftVbUsq6m)|5o%C=Z(t=l#=97G{+)5BR^N z-Hd7VVE76K{OKkVcc|+00c{}^b@~NU5}B*hl{j^gIK-6@VrBV)b|a-#7IFag%0^?0 zit-<7e_cWOGo~amM|tnTS^)|DH7pNI)Y~{xh9U>y>YcwP3(q)K5}hZSlVODaL%V^} zBD@c)fCblX)!_LH31chq{<1Ay8cLyrNcEKtzn%$c{FO3r>)7LRjdd`mZz6Z_;j%v`ELgZxm@hHq(>H_9c=m zc)tq&>$O=`;D4N&iwyaPv>PEU_karb7qppF z;Qm8X65m^JUo()E!%ef|eew`JsN(n^rXYMMI4AqxOi5&p6o{elKAcg^dW7L)7sp5=!CdnRkEf#RM3uUe_jFY zkSU4pUeIdVB5Q#bIUr=G3R*?`^9pDsQxf03pmnuH)&ecEjenO4+MBdLuYmS?Qxchj zcE^n49pb}<*p8gEFY(Ct5pChq!gt4vBOijgU|zDMyH$KYq|Kj#?=P8>$Q<86?54(K z%s*84j9bCMh1mAt!Ol@$>!=?|ibWi>-)6+Wq}?285#LVRVLf>$r>E)36AcoEafrwf z#d}ojzo^Zrg8dgvNqlc&KW8wjxruO#B70r$QUSe2`|}E*|MC9=NXTG_t33paQ;WdwF>S;t%^g$`J7v1WFny7wA$jDo%&R( zSgS?@f*S8Jh$pn0AT5a7a!nxvawjDj$TlO`-?W7cwkLx$4)0Zwen^`~1?i)vBr-=@ zdnAo4-D$tgs5i8mBQ5Hp>qhb?_o;x^v_G!^S}`S&IiMSd+PKU}N}L|^4kV5C)ZWz? zqPJ)Zm=>ZepPio zqRqadIv+A6k-6$@cuZ&rhUSPe+#l`mJ!>(*f2l2AT7YpoG4V&kZcoVWRU!M5HbV+z zUo<6=Ib^p!Sev92H)yU%VmpD1CbwF3!;T#VFGYQZasKwKW;vU4#o8n#t$}BemWF-t zg6D8jE8Fk;R7B@!^QItrjVX!D5nc9JampI4)bi(X1!b)3#mX3|JGDhg+u@d>!%f*f zFES`EP(kU_Wa6-n$ z566S##l>*E>5tNVV;-XAJ48{dI4cynH@3B#EiLeef^1(&ZqDls<)&#VrTWI?PV|f) zQ1zyv&Ap;GHB%Cq>&>diicZTaHG5gtgmT98?b;%x#dH-ctkbef%|t6ZqyqF7ZB`V3 zx~3$)a{>BwZIRLfB&=;@>(v1jl#gnEUIFDJrX;>|LHTQKk+eXGY;8HHg7TNzpI1Qn zk|~MIL0Nu`qN;QVR>GF`cp)3HMLV)`=vUCKiZVA+6!ee^%zSM|6kz6{WT{mY%gp_J=XJ6hqWdxfJD}YdY5Ung&O5v)DveFjh;?Va)H)DG!u_y{`<>c+ zD!BKVlE@tQwP)dM9;nq?w8l}lCR{9O6rxUMauvq#jJ9BDF%{84B zF|?w+NGWB6eoM&11m>zaSNuq>>XxVQu}j^)@>BSUJG3wHkC*e0r}>BOE#HwYr=(9E zQ=$BIQyju{@Q<33$R5fi4>|es=W+G5+iyi-jFd1g|06wwTr~>_SLDh*%L;rCsqp>1 zY1@bJ{f#M!ID9(I;n{=J^RKyv8-LFh^56|fgk01O&Kw?%X*$ z6C9GUv)HNTi?s&r((M*EdyNp@B;Kd}EycRuE=crj%)WB3)--L*^~Wdyr=6$mw52ND zbe}MH>{W@I_#thvlgC)4i8~u|rGPseI3IGSH2qO+9+aj(WJ)6JNJEU?Y-4nXP{`%T&nzjOP}9Z@@-_9e!!sx5q447cQ3)Lp^7?F8K8%E?Lv zS**;xEx1=OJ+IA^f@#r|MCO<-NU~d<;3=c>!`kepMP*@b+_RfXv%f+6qe`=POi9F> z-C$S?M@F`#1|&i*Hk%4UTwr9pe!7uy_Ko?Xv+l(fJlXTneD?B0FdTkDTcBi(BEe=F z4u41cTZ$rm%#=h3;cp2Nvp#O_I$o-tZ&H#4Zj^8@z$7D**G9?$yb`{b=aFjZ0-Tgm zYl_YqeYjbvMGdqO1-1)c)ozxQD0X(|DkK*rV)$~wn-+4}SdyYe#C9MTEi2YYF2#oY zgEoT-^nYtgA`ZQQ0q}b7r{`Q%zjhT=%5`w6i96nKF;vE~pUY1Lz4-cFQK$A2N-zs7@f)XCSBKw_;RaTVR$rN^qR zMy&w{H~?!q zgVoEMHJ-n1@>kqX2*1#^jzjCKa1dj8P1Dq4w4TtGF)dn~f))4)=FzTrJa#4Iu2YeF zi8f0Laxda$?=_#EeXn85kaeWtzO0RPLrg9#r;fd5kebG}g41%4CG1qDFz<*CU#Z_> ztbS0tVbWsNmvii~kbm3fpT;oaAGy9B|1xw*-R_b$j|zBCnvw{+-6Z4RtJ|4Q^w;^l z&auX8wSqN)9io+3kLyu~5&SROqNPP}UC`Scn|QMYZ;_rhoY2>PTALvSuAej|5r->u zGG-o&JuY4s7+UV}1|&i*X5UtOG5y=Uty*vk$v0^Pz6mRkw^=J{8dAQ)YV}8~?!UPv z5MlW|XTP3Jm)0A$=2GTO6bpXD&HV&3 zRH7)>T&#E$yKnAFi&vfb=Mb@e0w<(+Ygl63@gF7`>Y`~Qjp0Qx$dtbd<0poPhrayt z>%1wq$EW!Gq8@4L(sNljo&J+nm)3~1AvBG>Q_39=M}@gbUiC{%n^V|uwP^Tos@CC`@s))X?!yH_I9ePFR=7b39s`b*&eIIiI^oda#ZkpHq`_dn zRl8}@f_+zRjAg?F%Et6itP6_6M@Tjsve^Za4Tn?<+2~wxl1;WINpAa(XmhOClOHxE z5!VNuPU3M9j#F&a3`m4r%rH)w&h`4w)GMA|l3R0#X7RXg{A|O)y--3~3)WdI940tC zdatq$eq0-lv^r?#n!Y4~BSbfkBb`!6BKGmZ&2A|!(mZt_`)zGD6~+7SrX(_l{1#3% zZMLw>30t?-y4qi1eu2&^SZq1&oCMWB$QI}6Q5y3TD5#qzfLN{Q6iq}eQeiXcx znUYA2mvKaU|7jOX3}ji+9NsMi52beUQY*6vv7iy$n89Tf|3gb!@G6oK9P@t&c)mM=K-Mf7C%M zVto`_s+Lkat@Qi-{m#wZ4bYMZW+wN1e)sXazu&p%{LZ=Oo_iO4K=XYR`Pha#%94NC zzoag`@R#>H50Cc3FLVZr@_fZl*6Vk?anb2&?=69}$~ms)Ui!n`A*|`Q;fGxOiwCO0 z4GyaYs>1CNY7so1J;DKpD9^I4P$Ce<{Dl0Hm$P0=(QW#rYLqoHi3_hcNHGZLnu^=6n-=aoyS>G&QGpi3p*M$?Y=wU{R z`hU1yt|$7PRY!ZdX7ye!Ow<#H8q;z4#c4QXK5f>R=TPoNJzPD1tlzdCa#x3N zarl2b>qS1$Z)cD8B9C`3a?~}p_wGA)c*o;C%wwRx_uH~VuH+{JRnhlK4xlA=8*^CglParDlc7ustYJ*btl>;btPxCVtdUH- z@^*=LuuP@(7FwLL=B-QHm9!}t3g}IUm}Hk>YCNZ~gaabCPw*DpoVBjT+b!M#Z^4GF zb-Q9A@MnsHfNv@e1|C!_0#0g>!-|2+6-$6o#Zus1iXPxr#WLVqisitzmt6o&&hD~ z=PJ$uUZHp@aJ}Me;3JBs0bf!)9k@?%4sa;G_6B8hfeRGp0oN+d2i~N(0QdvNGk`l3 zzXtqN@l4v|WtHh5kqz>o_VE{)djhgWptvmK{`N)t-c=23lCPmn*VrUu{sd+@}KT`LhOP z{T@f4_R_hc2Q6Z*v$Kt}gW>Gra_oTzWKa`Qq_ejw(xt}@it@Kqpi6~A!RnXo_+nc$w`baRGcoZ(JxFP1YrrN}ev zQsfy5DrEQGtBj}$L89~cxL>VLKOhrb} zixn9`yA>HhA5dfj-LA+8`k^8tXhp3kV+5V8$OyVdaRTsq#fiX&6ej_Hr+6}Oui|82 zb)6`i0-U4B2zt39Bj`6389^UbWCY!<$Ou|IR5UPxPFG|EZB}FiO)D~j-mAz6`kW#o z=zEHcpq^o(j1hE(A|vQ(MMlsrMMO}y#LY1!t^_6S58PfNCGInWQY}Uf*Y#f3x#cRb zEJhVs7VlDIS=_3~viO!F%VOaO(UDi$=bCeuvOaNpsn#b&sZWYL!wUw*qK_Sc7L6Mz zhw|Ku49ZSHN8nC(t3a!sGAKLkas=*FFiJEPF`ihH9bqtpH7d}PZi9081CGGMUr~Ye z?GuCC$<6F}#$gW}zH~Is4S+Riwk)49a;vQ-MAX8zToIsNt07szA5e4a)kB zjzGJ9;s|tWj|%jpbgXEiy^V^D4;LviK6EM~KDa*2H$IfXhn?JBCO+&}qz|LV$y)kw zra`gE?+CPMgCo$Ur&OSKyAWM3LHB z6{+nuMf&)pB7J<-pq%4?BhZ_MdQnL4FEA)OU8e%G;(mj&ew!ok;13*uSux}!(Ny9^ zK~CEqr%N!)SYH9_U!->xV*Or&qO`h1*GtV`q5}O27?hndj=(cN;RrnQtB$}bbifgc zAdD>)#f9Et2umEH1VXDLltQ@85j+rnr2;c^w<0sM*dv;lnbQ@SWzCArrL-c8_PvTM z+RrJn>3dI+O`oSsl(FfXp~$9hwIZ9oE=4wd_bamL+os5-?*m0NeQ9&M)EaJcZzq^?r4cc&?Om!X=}x+RqCC%#tFksm^U46 zK@YxdNeiC9WnKB~P$bp~xhRZDJ1IbUFcgUR6Z*D)lU;(D%ko`=CYgiQWLJfgc&eGR zPmk^K;jv{TmXi11)ma#y>;kx+bExh8E73dG}7Y1#8o%s#~Ftc{@j46*LOswV4B zE^UuEzcR5hjtr8 zOQusIb9Sm|_a}YM956kc<3AbkA|p;@#D|QykP#0u;y_01&xrXMu|6Zl_lWI1VtPg_ z?-9c@Vt0?2-6K|K(!FJ=Fy5VGyY+i{9${yIw!*9cXW#U{@sL_f+ph9O{MUB-Fg;0j2w{s|tj8m#^#|c2S-WVtl#o>RMpR}TXT8tG2V8uJ#Vy$l z*_KmOKoB$ef> zWvBXPe_JSfIC~G|Xv{ORugP|uc?OneB>MsQs!A;CUINK?llf@I{)!V=`@}>@wAC04 z=|7adFx%WxGPo_04zZlrMFG6XMzo6&-#YvnFIaRpX$?fe8i<575D9A_64pQ@tbs^a z1Cg)>B4G_g!a9e9l?@52*CxCO%^H8+qNg|W?_Sil;&2DDE}6B%G%6*E3wjt7xuy(Vcb4o?R{8_26uS01@=k*EeHq& zd~sx<^fN|clUxz>G-OLM-P#)Nf_>lV3YCTe zVu4efbi`jORYxj}4NkDFI+$7n z$xNmL0nC631+xg}Z2Q={tUv)5H4{IcKg?>Pw0POrEo=9EbfL8plGV(`DlE|X2qhA+ zgf9|9*}Y=m?_>F-?>Up#VkRc;#1bj@9j#f2@olKXj0kz=om{*34P#!}?(gq;uya;l z^Q_e{(!Tt-*fraj#R!#eR{XIoYSyCt8?KsIcI!KR%tBT(p5-6(`Snfapo5=Y(lP4o zz6bRjPhU@zdfHzDl#^8MD|P=9xA1#)B6+xwSC!`h=~0 zFCq#gThF|MrcJzbgC4YK|DU(cdHa&}eSE<|Y-{rmicWge9JKHA9dGU|I6?+oc5G{; zJ{vLav*ph-YWmX;&NuGYy?e@CtDATB@x+wjE03*F?o&@=82chH%r5E;)XQIsRxf@d zH&7RCTzmb-QD5&3)a@;~F*-lo9-pvws#=@2M;+4K;69W0h-fS9T5QzbhcBK8quZ0} zhGz|U$v+YfNEdwOx@d5QmBw!^q9db?n9MkxZYA*r6?=H@u9knDKN(1b<0s zz}=;Eqh==8wN3BpK=;g}avZ~ScA2sj4?(D4XKE(f<|0!D}22@)M2@6!_jrPXN)CyyJL^Z-NH^|5Mwqu+iTdUBoN;J)aj-AR^OSf z(zHIn{{5_(>8d_`PW?}vI(6z))yw8TKX1-8bLhYJnno>Os9YL!D-(r!wNiE~%}#sO z!B(M^J0d-LeW(BBoy(oR_FSh{>zr;c$yKw>y6ff&^-lYNp@$o8y)jhI7u`z3tqh%a zoh+!151kS}-IK0Uuh8FkYv`n#ZMN#}>1xw$42_J^mxcyCa*a-VL8G1(RPQpT3L@ER zWvnpXX)pK0^Z#COhzkOF7FQ?SdcBZy8$!HAC8sjpa>iZAxS}>SPVLbcX*8!wF0m%s zs?%PQZRrx?8I@3e9SLia+wO|B-C9T&v}$z)53pywC*o!9q1L{M1@Q z#Qvp0`W0;|eQ31qpefM$Wv2ng(PZP*>UhZ=aVkz}s);X-6mrDfGBjx28L2oEXx>hH z|C#EfTX%ET=#+J@W994a*#2!kzg{bd21mPWd(i{jEY?9laPN@wqK;2)tcqr@6a7{B zYS|sBK{-YW6*Nn>^K5&4tKp8w%nHEJlNVlh@l0QR++T?+4iQIQ*ld; z5uyN@nhi8WZ6w=h&=(u6TCMIj8Y7e33mcs$AYFN?(L}c^I|b;b`fO~*Pw?c9KRI+dK|^=QlQ(!j-OHS<+8__|Zodef~~$BX!xw;vF;phHSko2nPa z^U(HryKmdQ3y*|MrRsRK(_TbB25S}Q<|_A6w$#c|8#mGB;OhxkoOL^Q4i;)tmC=!G z-pyXDRiXZ9oW+BO&yAdeDuUm|=ku-dXk#!p3U=6}blmp!&<9X)&{M0~5ubJ$jSo|V zhFK=K7;`2Hunt0j#Dh_%;ZhGm1FwWyvjy<%2U)-+`KekJJq9fTB1-_cOBxCT0w6aK zyyaG@iB6Io0Y0N{oxb*LdjmE2h+A<+p}NFbr6J4$iMSN~u38xxb3q2yC08w#9H>DP zTECDmFN{qAu|{jO0lF{aVRLBbN!Yqwm3T&vLD)mF3CYQp-#t^nV8jeKI~o5 zUheg%kt|HUOO3bLf5P?M5#N3Rym`bpcoSP?_khFHzK3Z z$!}@6O+ozc;@> z|DybT`TO$+@(1&W^2eXcpFo8N@c(fBWd6Zt^Jnr8j||_K ze+eo$`O$nfpUb=Lg_m4LDnI^gKA$hN*UPOo(r`=g4ak&ybdU=MtU)JNu2u@9^Mkp1 zwU(<+R`9gxjF03oYQqvkMXg+6Vi4xB#D7ZV_6+5oj)w$T=Vc1~WSSZy^kwmR`L*C7 zdW>hCb1>lL0u+>NJK3#8BQ$OnU~*rNKc2i`KeK1x^39!L>y@3eo&05b^rZUeR{7DZ zI_;(0Pds;IsfRHQWcYM(luN1ak~&l8D-{~B#VgzzdxJ!XVVFw}khXWAA6~U^xdH zM;*hajl)OS09RYmb(yS+vWNlV}4NW-4h(W&_R*?6(8MxiO~C4LY>~zH$zgPN7cV zM9yaqiAIO{=!x}eC;#*9MTe`|mhh)Vz5KgQ!#?Br!t>SadAQE}dLKVrbQBnzfbA8U zKn6Y60{=hUd3h)Q{`R6KIn<;#-h;)|d}KWM2?m1$qtH`$k*<}6tI%*e&pz;N{P)~* z&o(=w`CozeF23~s_NqqHX_D~p)Wopr(_Ds;#wm?U58d{`v+hK}owTZB*16y{YZ4Bg zb=GY-Wq1i_!}D%7Unx8Vmqxet(smcjTqc|%ZFt68UyfI)M!Stw13VF(f&y?f#LQfUCIO8|(l!)V-s}7Y~t?+8q9E@pTF{DjyXi#lg&K+|w>w>yr)C|Sj zL=)mMgi^@-!5^+8oP(Jgz<4%zcNNtsS82RO?11J`Vc8AMdZ7$etu|@$f{FeZd>c&C zMjAB=zdRQZStk4|o(V&5{O9CV7e^V>_L@8>GqC6nIwwIAu8;o!X?L`TqVa?YNiQekw6wRO+nokQf|*x>#m;uwt~&&i<<$gwNT zdHfG_+K-tNu1)0-J3u>8bWtVk2CcHv|K7+b!f+IL3yPLs#6S~9Ey7mClqG184S-IB zj>wnBRAB_M4vLRF(Fy1*f%`e)!q2XbOS&PRqB};Pt`Uf%x1)udBbacG%lfdQ27Pc4 z{o=q0j7e=OL~dCH0Z~K?IZAl?GX` zi&LHFM7-y?$1OPB2aspa>aSN>337aO()vgfx~QHc}sf z$OtiqPnR~(pv|XR=wEV6uaO>)IPAW(=RS4V4V!zNT*ZqYu$5ix+aqDqt}QyS!iz+t znQRDknYn0v%!I@&#BBxgEEjVkiU0`%SA3GXF2T4=6!>QN*!_z{^n6Zx9z7Q}dWopP zld$5S;yv*Xy}KqXk%GZNvG<~gyonD6B}>JNUU&_I15qV{%%v|Ql+dkofC(>ros01;dP?5QzY1WK54m9sMlG&GynS!Yqfk{9=mb%h9(Hv!XE~02; zvXG-L8NSQ1iZKHLJetC?D-$+0*AT&Cii?9OAN^H=I2J6j^@3Ph=)!`_Vf$gp5PCv5 zC5qK*@&05TF_p?V4SX2Lmmq*!cp9%brGN#B1_dT`r3eEE?e@nVzly< zB;jxRCHyL!MG=g|s)883$s_e^@G1_w28cxF3bZKjtY1t~RsZ)kN804>?Gy2O4I(Q| zqJggeT<{i`krIutUmQ}W^*-)xlU^C7Mi;|!isuG&m{IyV%o;)xv?ke+%K1BFKm90Q z&_%s2N5s`KrYp=EG0BTQK?C@`l_HUzj}(^iVhmX=04JtOt0S>txRIDNEk9I(YeaDy z`UoZhgd2s4ZV-$hb?MRp#cwm+`UBW8(XE~YNa#G+065L0~U>HQ#J?v2OAUP2;WOJ!x;B!3R(xgz6G~6Di6m!#q6c?*uTu5 zGsmCDSTOxg%|#sR$fa5zLaMOi{nj%}Ec{o;Dn-3w@xwtZOqSq=^$k5fxbxQgp1k>) z%YFSXef+``7cM+8G(KSMxNza-?K}GKv4)1Mo3~qK*Qqo_2yrrxX_N(*kK&LXu_c$6 zEJ3|Pa2iK3d$1<+Sn6{SZLCdMEUbkk!YtwA<_4`t3#F1Z>RL4{PlE;pCMy_f@Qc&5 zCY_)r(FB&QlBiR$#_P_gs7*MEaH1!KRV%pYPZWw2MS>8&S=rH~PeGq^R-s~Xz4 zT-=a{!036*|)>GHC9?Y-qGJTxU+vpzaV?r!kUrmsnZPfB8nI) zdDxWw3U!tm{k+lifbtvj@h8F-V#$wITUc@SFyKO9Y4!2j)DXTlcz6yX;8SS!?YL0c z(JwcNXmAjsR|dr&eT)fK>TBKnWsizq<%`&5ZFQV2R&Yc=xx zrN?(a5q_VlDNRog=|AOt5~S*y-Zj)*e>#s1Ae3rBVQ`lcogd(ecLWn zwndapszoLuv^Ikc7)Q77wgJoD<@Gf|+OyeGP1xE_D;TYLWM?_0aM^qjE(^U6&N-q5 zMNF4RXpK*DKHBWPaEJgo3I5pKzIMXZl!Yp_>PM9;* zY5uuQOrX6q778}zyjc;NUv^NimK=zDy>Mn|!A`8#L7IwNsrrdn`L_iX`HSedc?|88 z|Aa4wu?68&>>Z)G4xWd86k;zNSmqJ?aX%|4o2wKXgV|EGMX3gLcMz>P{bg34r^QTT$>-a5fToHPqXm(7) z1(9xV8b~3EFqR!i8<6lq+ai>4?FB!m?bjn5B#Ul){y9&^*F<0(a8f?4!N zznbvL(K{;*G;9?Ch|NbIK6%nQec;p)>)s=$4xBi7|AE7Y&mK8**uw{o;nkBT?=KZ7_bq%a@8%~-DT|L#6FrK5go-ZVNWVx+@`88#1SyZu zE#b{`^sT*x3&?}!-$cp}vUpnMiWj%@)(xo2CHV1(kBN_L=K9TQt>jL)m>%;wHp~p- zdKBDQ#}&F~5u(haDI9E$8(dqdayW4)BohMq5iA#oTqZcz$fbeL zK?{+c$KC+(g&87Cc?&J`Vl@;_mS4j{fEv9|cvkGS+fo*ZEofV}1Qe_bkk|_^KYkLb zR2eviwa<%BJk!EzVqK&wx{b?0eX060tqL-+utrb)EUa;hg%}H2Y!O2cBR+-{1Z6@R zn3OfE7szcwm>;pbaf*FCa{(5?ouv6E64#!&fL=;IUxOlhm!I)GhQ{UaLl=)l z`8*TQrdC1Xk-ti@pLBg7TH3%Q4T>8)3@P4%1>f8iG-jtE0NlU;N>fiiv_-}FXY-7+ zrve(v5HA>nKvnxl>-590S%PHs}JK0?>~Dn8h1u!YEVFdNiQ7T-W)!zyOasg z9Lc*5>@iKD1;y2J4a`|E!5Y4%yTnf@kwUldEa(Qm<2Nj9~u&jY;UX9LKsEmm$+sG25&yxrWBaHi(c&mNW zgYa1(q>FwTL;zXEyoJT(rwR>9K0Lg)>CL98_j7!RAJ zy~tlHW6;<81S${@>%WB(Oi;_SOY}an3%TXUI3NzXaOZye`?uZtl5l`nE^2w+8f7oR zN?Zzy*NS3V!V7#Ga(nzBEioXz!Ks*0x5uxvOSfg@+pkW)MX-!1Vh9s zuUMofz`2qa$(MO$VrWP2s8q25BAj)slZ#`f9{Ef8DgQgg6ErS8APr8%Vf+!!%ZexQ zXE2Q_p2A;21h;rP^4Ej-t9XX~@uncfhv<*9(i2J!_6;>92nW9-VL135NyNeLkB1om z7lL2qd~`NhqrSmfO?w6qW2jOJ2PKip`6I|L9rfmnBNtud*Dv4+WH_{}2`gaPA2AB3 zJ)-@-38oBrer2RqY9W5nwLZ8mEEj`N3CZr5)+`puu9)T&0#cMM9gtlo7InyL#cnL8 zei1`O3Eo6dzEPCt$Vc`D8Eo01H1BW#;jIR86;4+PYp6tWh`^wI!YXUUcX*KmK8~0# zOU3++Fge@X*&s2xMiB6lhj`zERJvLcLn-8MJ7p~JL&_e-n3qtHLoMbO6fRexJg%0r z@Y0qGwMwte8u^C;9GFMn#x$m|b}Eui0`=hbleTB%9}QmiScHcR_YfPI*o6%P7c0ce zLIf{uE_ihlH>;nLBxcNFMM?sPn_k{86488S9mQ|~f{b(uIb8Ehr+v-kj_ZF6Y5Kvh z4+g)sm%-MIx^Pasth_#?dIqWc3kF3g^2_~aoRp75H+VXhc$umNC#7u*B#`KR-*3O;qc^r(RUHbZjROhzJQU*!K+3;5|04n6;01q*jGX}FAELttQnZkNxtt*t3*KE8P?Rx(^oAd%Ekk@G&-(`IYYof4x2Y z^-d%?h?QcI`5r#sjbDR2Whv3VE#)EmN&Uem_JUZLeBF6%YqXd{GhBv1tGt%r854qu zo#AW!?Kl2*yFmz0LB@y+WxU4UXb=__DBS>_F3SAfxdAn7z8#+1Uf1;zUMEKV&0PgW zXTsrGBC+9ZBMQDm)|H;G?|RFd3!|j_`?QZE<}r4Nf!ll&|F0Xqv(;YbdEjK+04A@P zW6^a@YxM3xAWnb2O(Vc8X0e{sa6eQvbh zk#hOb+n&X4RU`|N1I1fYEkCF`IXOAV5>XUz2)^AAYBq*oq3DeurxVKDUREE=N;D9L z<~_8{Dm=%*g4-MuG{`X1`&2vf{tB_(2;&JHQ?f_JNzpqfosbe>MlO*bLJdK)UKrr$ z(1jz9b za|CQPqinkoKi#g?UW93AuzRKWc}%^I9r%mm@7X;6C%<0&-@#uO`kpU-!vE{$yWji4 z;-~$;ZrSvaTJcYVzb4-H(cRZK6`hV^C>K}fn_~-s#eZTqg zFBbp8|EvEa-}lDipZI_6?Ejao#g7Mn?HoQ*{Hx%vf%EqiKNkGe+%#7F%iym|Kl`=f zr~JQe{hvF(RQycv*N1n#x%k=Oua7+R4s3TuS{SV%RI(JD%D)~Ni{A4(DZPq>;lX|Mtc)!sO@v43A$hYym+bIEq+jI%Z6_6@ z8|s8)vBik;7~!m1DnloS zh%KSDF9=Z^qD#I&L?wIk0uhyr=?g?u8Yx~NqS7$<0uhzQ_!o$%<_5v}p6tieCqiCNbHYDp=~1M(pk#5d@{2cympU!%5_Ygg zi5;rNL*iRvPEb6I-}w|X!6p_Tj^hdB8^(|J){|INN3t$5X1(J91}sDq3R&xjpEyH@ zF9bp+s9$^tgo_UTfw;x9_(^Y{$4`9p5&rQQ|9CO~xWGT26hDUNI<4YM@T)Hc&A?}n zV!JrG4WS3iv5sOlqII-d$FWa2>om?ZLW1RKw@IlKQM{7A#PZa@k(oTpyW}CCeg=8g zNXJZ^;xiMXyvJ2wK7YWNPn=E^X?JqSXW;X?v$S`WkH3KORwXrx@;hBc=KDvD`JUg! zH+k1H@V)kAbsUQ*(DG34qYTH^U{2p`%qh+ci?lE~$1`x+-{YhXu?hG#DamfUwx3^+)#zy=xG=z>JYGuYmfE!1GU4tX8Z7OnjHO-DB`vwoGf2DA843;+i}$R>9RIsE$5GozvF;i8eDXlz^BM=Nqu+PVK6ie2&#ry@hW6qg z3mdXX(H^c8MC2DI20`^6Y?P`4)@kw{rlR7e{V{H}`PlHL=Ve4&jpj8owAwwM$LAeX zHDDb_o-h_6$4UMJ^*UnqVn8orAA@y9gt6gMiPX0)UA-boEbm%vd5fNxcx{(n)B48@ zlK1&LZY_W5llSUC_>0&9Sf#ChIMe}~!6Pnj^4D0VZQ3$LF{BCG43gaIJLiNM)|^M3 zQEcyl3ZKurIHgwLhm>|c@C-IIa-eBcx+^2a65pmRv5(?4T`1B@)eN$)6DT4pT1a`U zE|N*2GR0S6UhmW9HHr?cp3T7FVwyA3*)m=IMS9M>oVqezTJ0q`KXtro&MPLGJoTeF zm{=_-((FYnUKVM7AQmo*|t7O5|Y#mFM*h**d$5&~iovMA)oaU^s;BYCd04$IT- zrK7D<$>sfA!DPK_{wn6@i$yXNHagJ}f^pN_KBV4omIhYMy&%A9Z#xhrgE-iFTg&%& znVDG}EOz&*CZ^J{^MBcr6wgpwiP`Z}XahYvPwTC)PRJF(N;s1C##(ix(ou^cq98J< z#MB?tdoqy45&YvRZk5_we6^(gRgXk5ATcIL6kmcLv8`Qndz2I}_A03LDiEzY$j;#j zIETJ(IBfHtZ8;)S8mCOfd^u_ebYJFXdMy}A1)TO!M#NTZa&_(Mm<8y9CWo${=rH?jz zw1^x0o7xg4ZSduu738!QY?8rP^+ZQBbexs@hBhNgQ-9r*MA$99YD!`RFLH}A;I2kU z+!G|^A|+dU)A=fe%i&U6xXr+|PPfVo4kDy#H=VIIi)$}PYZ$dMP3G5jY?`Lp*yiT* zt90IZYUkal&6%Q6J4{J*V`-ocp+>PcW{z`2T6EV}tyh57|rKMkHN+Rq|ml>?to&{}wln#)>NmJby(cBju$fv2tV7*gY z#H3(t&P;+wyU@Tv2DB>W+Bm3L+R1zFD=(7L zskq`y3GB(hR&{Z1LlnRS^%=;IX^Wf`$ZeU1$Fq&du+2QWTER@sch6qkQYcX2aEu0#aDYkQbkD_T2-mQo)JBm*fRT@;+^$llHrt;GYO1 z0Iu!2owb=`yIRL#LBz2H=mHhcd$k!;0KLnUMCyRv+!Ii&ui}C@$8ww5D5S|=VNi40 zf+q#l>H#VprQ+Dcw1nzH6;(%@GX>QrO-ZDV>ir@zV7Zs*5+v{9qZ7!_s5c!XXq<9z zRLUi>E;k|01G2E=lh>BLv$48t@RgHeZFo+*!IEmjfsBYP*uKQVB`Ww5)0th6SUhBj zZC~QxvI$vT7OCp;N^RB^b$Nv;iNxw+JY8SUF-K1^;?oFv)L~@br7c@hWH)$GNDsm> zKgB`n-HTPY-l@%x0@phPiK{rDUw01M>2aiQ83$u=AEgtzBhFBK6^7tr+9D@~U_E9U zW!GlTxGxcBG6Hck{Xf!XNWtJErX<2H^9Ky$Y|r$s69L!+-DbLq4BXeWg-i#46YDH@aCqnDPF3AbX=Pm1*9XUBvJ=yWssJA23qTM4VX$P!&B21Cn-Fu zGG(NRi;NMRmqYwW9cY<~O<9{01)HKNiNx3#_?3OUZ|RnEaKe#fkK9e#zJyZ}{o76H zYM&C$SEooqv{xm3hmm|<2+s*%3vm1q678bdUpQ1GTew(ga+tdl*;N%Sig;SlBL3FT zm^SpA@nU;{w|}jD-28?a|7;=e##6d18JCYg@Uc%V^46}s=Ae|U{5Dx2B%F!dGOnOA z7$(*H-I<2K8i6vx+77YC;xDkqv6vKpiK5}>6wDo7uJ(=BnYMSB2lr#F9=-D5W_?X= z?Su7kr&4&Dv>ItDmN-pC#-SZsgR%KzZF!Sovo3?fg*?7Q`V61)BDH>;1OHKNeiYmE zzf4Jlw&{PGk_bIO)&d=+_XG*Kn85=yu#hut?8HHId?I>8eR?mf2lZK*zOCJCNtJ0{ zrjWzYwW0E?P`mCowZE>Y&Nl>!t1|!g2CI3L$`~Ds_z}10*2Uy@d;5*ygmq;jw-;Q% zdg_zAqeG1UPzKSr#kUHAI(u)uM(W3l#do27itna>-h+Q0s6TO|HwJioinQFtmVePu zR4(bcESydkNvbYuMZP7P#@=lQ%*fgkXL@clZJN;ZZ!skicKS_%#H;8& zT1TF0iS=blnm&iEAan&y9UCdJX%X@FEUnE`EQn4|lvSv0NQ1$CT)SzKf_-~tjAcWt z)y7f#LN-dF#`+6x_JwRVjxcLbDwvRsqDdrML?2?34<6IzSg|J$o090(2c1shaS;wh zY}E`%gj~!p6qzRe4xFRwspxn^;kk&tNSODY$CX6l(jkgcoX6D(F+PpntE_` zEH@mN02~l|W<`E{WYp+YiIMwrZQ+t4w;>aOi(rf<@6R1xt>X1(+WaVZeaw_ZV!Vt! z|9j55$lWA!i{`Ls+1jm>7m%YE311~ShF@^}Z1miPiGn*B?U+|YiJ|{b?RH8E{XLnw zC#MLCv7F=)cyiLOo}9LHyo4>pl-z= zMrBD<8Tz;x!gbn=E8uT5B@r5jb*3ajkC0Vx7Cb@&5+N6yy>37vY>OvNGYyFgOsvHL zQreL^WBZ%*Tq$_~Elvje6tddL$6HL2I(DDJ_s8v1c!~B`6|H|lkmzahdhb&(Zt3d| z(x&xnS)j+-_#fKHCDq1* zu=~Eu*r!S|bC{B3#;zKY89&6qlnr|lhgaAnGufCxX35UP#4i6`n|DQV|BdnPkn0NuuHZnCLnO&NnDz(#03AQugdUUzjStH8H%&iz9ECFvWQbLe~>us3hpf* zzBXOWN}`Wkf_a z8jcnG$^9vAD0q=JPYP=Hn36~xwHt)D>{^^4Ls55ed}tFoj%!V$_H~B3jNq8Is7VpL zk$q;|i7gU5z$>x_vC`-B8o@DU17cR2GX=+&n36~x#|0;yDP`B8|B#XS5pCv^BC{}4 z;;%;_h}mKItTqb@1V3y_B6SE>(;?F|nu|lGYnUTMyH89x1M?njk&*(lCNn1Sgxx3L zDS}bVj>7*}n;QkA|87bmb&R$h%HxDx7tW>>77$|=T@qg|Jl(zWo#X{Z@qcRzofO4P z26tkDYq*P2m_UqVA=eRDWA+vPsWy8GtbbxkB6V1=e~{6mi36=~MBt6~!s)9rl>ebE zXi_L|$asj6yFYE@=i1?zZG``>&6on>znPLq9mEx9-SJilo3fvdncH<2GEg@yOUu(= ziG`#HHQ#xMI;<^R(mu6; z_b6PlyO@h07PD*bfHp@8VlOf!(T$ivd?pO>@O<5XM99SqAs*9-{rVGSOywuYEZ5vR zT%$5hCGDD3^r*vH)6#C>(pf#>`kgq|J~5*sq(C zNFA`{j}#g$rxfcHb(JwZf2%D@Qg~KmCL+8@sKiV-{#u(41(m-tC6PKRTjY&Jme`3z z+b3ci*J*1qVvCoj<;7l~k@>W{_>}OAnQJW2=1ReDt|^Js@$2e2a@Iw=tBf%k)D|Ua zuj<<1vV1sZa&fEn*A*=KOi84U#cjboZQW~rln{>m{c3%zRGk!2nqgdT)fHCLf1MFM zr`;e)5#5y;ZQ;hixD`E03=bp{5gc4?E^pgiJ{&VsIit<00{3#sT!WT2 ze+s^JQxd7;yYna=``K)Xl&okEpWX}Ad698`vvwmS#ran3Gzz5XTBo1RTa7s%;it8k zRKWg8Qxd6zz2nh*A)B`xfPk0bI>kvvatZhgKFEHRLPd%1#PXpN=Mu zMv>Eaj}d)Fy9tsax-;Y0z6`Vi;mbgwYIc)@L{Qn?h}pS#Nt;In>6R&p)REqI5@{J6 zSQGaEH(F>b#36BFzJHYf&k}R4TSL|`r_M%EH>AND@Uz-YlT-t4&y+&h#IXf|Y{4}N zJc%UPVznV=pW>Uf`Bk*xr%g$8YlFd>WVo!!TZartgj~GXG#T34B=+g6dymbUvu=aV zP^R_LxXZqauM9SXeOeo$c&#J}V>g7oCT>I6C$+z-=+-9$34;w`rfbm4BDd2-BE8Wf z?(W~v7BOMN_Pn@eURy`_Z9PL1_h;wKPxbz;^ zy+*+4Rf%D`apmmHNEH2VZ$#z~b4MdFoh0LN7r0yV;d2M} zE5`Y1t_!WW3XIetZK;xWsIHZhd2v@V-mm?21(yASL{Ez~>l-N+AFNiJI9$@HMLUG_ zoLkn_mMv+^F3I4gYL5p=?PIUl8r9}PY2gu55~1Q>FeMTC8;tW6`5P(O%dL+fE0WG> z$6dfRaXxUiUTrjNq`MSwf=;1YiE2uM`m8}eq1|dpHE3IAf_a7o1n>qeo$M;q!@1D#Vwgj zI@XH0l)#Hg$$7svQwn(RGbPato&_N(<9svS zN$eM?T1d{iVMl< z6dp=+w>sHN?2Q-Cyo=Oe0Rq#bHV^$AKl7A2*R% zt<9Q((h5@&VIN#7NX)vcb|abbbB`PzupH|sE;~XF11?8CP#Uk|TJwC@c@{HOW+?B{ z7B*pDYTw8wAxvP~ByLynn6;gNSqaCOedV`n^QGXp$CO0sIIcR9tyZe#DZ7CkU<7Ey z(zVrM#_CCJfs$giI)g42aAY?oa$yX3e%w6gacx! zGGwpN7BVShU9-{oajE4m)Bd`G(q&T;siV|)z#4Vwd|j(Wu^mKCDwG39rwn%OrHaIP zlE20XzC*hKk|NlT6&aSmjz~bE`7Q}0TN&g1{+u?83e0aaC6PMJd;NGSVcW#`cc>c7 zVd(sTL)AKRFQUmmUBMzV5)A(zX*Wz#`1j$CN4{7bl6jjmeaS?sW;T>8$SjfeQ_mlc z$%6lgHoJ-{{DCQn)Ky{C$wK8Ksewq-REJHam?8XEZGnJf+Jow>l!v=lH9+n z&58ogKbw;1hR0x%8O~FACSyP%)A=TqK1^>bHup1SsAbzDT|dwB!Jw+9o{fz`7! zKOzxhC#dg?o1peo;H=JwE8C+E!}GkhXi4GOASU)cZ-HwP&KMWO%@|*&&5wfCkC~E49j!Cofuka2 zs8qPj3{%_iR#M{aDTg^f9-9tlOWPo8z;1KJImR7W224>4`f zY$j}xKwBD=*Pf$w8ox#4JJsm0EpCyRG^7P_Q{7+F7DZ97Uoj<-x_YgDw9r6&24PWo z%ThFG)T0gq{W)#Xk^;RUGwJgtZ%C4jLClQxGur$pXno3*MCxcA@J!eMe1*E)$?koXQ73XL|^K<+HiV=b7sCMCDDeZW6wNXX2SSr!?G zfYqC110XW)LlPo~*@hw-%pwONjs4jI|;CuZQ{e)3@LyeHzkoeU^~tM zAL@~ObJS^f5*JC?XPsKUQLTL8Wk1!hI590c3g0r_TRVbx7|@2c@JRvPo+11wuE_V& zR+xL_coFw@!}fTW@Qv9AUeo4J!MAKmB6WOkJy5I@+B_;)zs{6I>R2z27~g1*?JDD*`5W4zBt>jR)YQs;a?IxK z4`}nDpz>>`BvMD^ChxkaakmO{R7WNQ7DRALRNaj`*k;DsjNBKr#Z8KwbV%twbbJ2Q_o#5>6Zu z3$gHV%GEkD4fD7Pwu+gKSlzHHPaUZ2^0=% zQeYc4C6PL8i^(x5MrX*5FqMwjx z;H)f!VJ6%P-3Eg=T@w%ejqpgEjWK*0?=g~Z)NX>LNbVF!Mr5GC8)UG38RAT3%nt82 zX!EEb{k$oO)RA6ws9KqD!&}3;7CF^o#_hMX1xkwBYTWWe9EEEmvZE2R<@z_YSyABm zfGLU8;aS-QPt-6fr3}vBX^WE-oK@Z6$k?n^4~zMNHYW-;pED(qIyN@~o1EA;0P9%6 zwFlF6q?sx+a!b~w<`H^zLTv}fn2D~m;&THQxd5I8Tf(IL)g0-gEyot zVA5_Ex{K4nwqAv8K$|56wtiC*sl&GJ93mxFL!7JYUt-!de8pE`xXx<}mK3h_h&PvA zY-hk}65={Nx?VMszq3Jw>LG1*6sS&{l1Lq@wMXmND?4SKb8AhvJc=y!-TR_LS!@jk zYD!zIq(H69j0wK%hR>G`z9heYql(jnHa7}R4O0@S#9?Fpspkxo5~@lZxCMwRuvIdxI&7)R9{moSGK55T}$fO24ZuPEwRsg(s$| zi)))zY<^3d69t>!FeQ;XHk-Pmg}BN`9ODr+fxW9ST3^%_FezG_l;~m1lIV3RYJaEA zk%HP6Oi6sVptgK{S}r}MW$%_pt1T)*OSHePAhghwMCu5w#^KgTBXcidkqmdjBPB4} zp<>FpC+*f2DQQo_osYzmE%PM#6W6N<4QcbDAT(e~B6WmRljdHcGuFBmy?R$;ydKpS zFezS}RmKOI)9Hqo< zMYx*2E)wv&yu#$K@ffyhLs|}efA29&Y(-u*uE8Z@f%Z!TAZtbrt7!8?{`0l~z#o8k3V04F!(akDGk7|Ei!RWjxiSHha zen4A99gOajF}g*?=%w0US1_6~C6PKtH+1cN8Y|!=A1ja5q{(XiVl=ujLtVCcZ`T$z zDONXj?{gBo+5xZ9h_zis@vYi?DJZ_hltk(%ZVF=YUXm0-RB=&{p4Aw;4{HmU6uZq~ zbl&3-0W0QEf^CNi+wW_$q`>yOrX*5_t&fu4aCVlsJdLh1;f%Lzwb2|Z;l$ry|9Z3^ zCV!1#{r|KZAStZN`J45&y57kP8FDa+Kec`>^3ElIxykbc9;`@ts~QqZQa!v ztS7VuObQl-7w?jN$Xiu}Uab9f1))bxNu-ESUvU^4q}tmK-~<&1`G`0Quu-F|KkVD} z#&>W3Zq(o3@0Wa;L9&jJ~)`zUaJDJkGnd1T)d1$b>yw#2

  • bd1p-cM@$Ums@b8FxVx-XcTZQF zZeu8$caWb{-#b{SO;tua#rJ^T1G~>`nG=zonH111stZmCBOPr#va`-s!TU9BrWEV-E2bn;2k-7PJmLoiYphyVHl6U1mjcmV z=MT28gb;)OIqgPD3jRGA&yxsxFr*BZG5t9i9qh>(A)N*#4{d|0a(qUcX+=3cWlAD- z<+$}>ItyP_2VVDW3|P{Ye$lS3^(MnVZ&O;%{;e54hXS50Un=SY?#Y1@!h%~DWw?h_ zxUbP>Qi1!wzGLCua;{m$kSvb%4{W;UPe$RLwkE^buia8fd-L_kA@TWCIECbH@(*!*8^kMM@Z)hPD_rX*5FXUn-# zVLaa~OYO^`_%fHJoEta&U}{NX&FfB{~JI>&&d))wsofSX=glE#BUo$%h<@#7l5U=bbe} z3$QTm&Wj(^W=VnL2Tb88UW8UiYcN=X@RnU_y$Zy7%qE&lG$*SVyUKLFPnC91smCez zuEuEmyte4c(P-j`o-Ph?n0iCL^T7~RIERBSx`JQwzTd#~$&daG*sadF%3RonK3~z0n7g@Kyr42<=-MSUJC8UU) z=35G+u*;GZlG4Sf+<#~@si@q)o03Q!Y;^lZb*yP+OSmq&Fa}Vgn8uVdbla|*oe7fS z!L-RN5Kp$mlZ4P*)s?M4>rF|dj?m^|w_b5e?E6%3^b&5#!zCAm=2R4^Y3eaz z$FxOE+S#tlTx1@F_Y>t<Xl++l5=6yy$=l1Lr7zQaKr!fPMlzn*gIS+_0@y@#uU zlSaGJ)3#6d8csv^Ut&~C+Ttfgbw?(QPssR|1^+nqK48Sp6d?s8}}-(_TfRl6yYB0HFY7=bLdEDJ3@vMfQ4wsc_F z!r_<;Vt!ehPX+b&n3Cv5-M|eFJ!uX}7?23Lc*4{_^WsmRyt?rx>#&=}6-PKUvRQZC z8N*J$6NH_Tqk{i7gZ5Oy++bFAVkJIO{kqLri0mRii6 zf900h@rsoOZm;&P@uZ&a*o1o+U0M_u%~+<*iqh(fO-Y13Yk?q9ob71Fs$)1+n(lx* z>N=Ev+hw@bVuok0wm=E>AMlJbNA{S{QQ-M5iVI}y(q=`0XV8>HVt9;$wp$Li3MG*^ z&b?KOfg_;iUSr0pUGAvd6qB-b_rO z=#RDeQIzJRg2Yucw>LjhXtbPC;b~e8#>CAc3hehw zy}`(Cx<0L0#Lby1@of{|G?S3{4sezN2aAU1#aZz6+WaZVt}!LijjVzB4J|mE1OpNw z7in4Wz5{fDHQis0g9DIba?WjvD~(&VaQ093q)W?^zs9O|LK}yqMx;L@?xPkG*dj%W zC9qpHT%5!43t{;q`W6#FKB~>4qHTu+iK}S(Z#fThf9Sy3^H$IZ0>$W{9JT!k-(lP? zYBx$!+_sC{>?~b09`G*kiw@ed`JW$W{`1-#DzLexB*N}F%E-^Q`QO?D-|&v{D7Fde zGp?`I7C9-d+j`+zcgMZ+kE58zEQP;Dn>z*5SDTVZ9aC=~k?*V>D^v=&LpA4i9q-Y* z8YB2VZ2^-a=x;ZoX){J~dmJ3%)>Pt9nsd#Mi&nf>n<)jjcbk$(9k~78_K7o7%{*e_ zhp?}}|I~Hm4r6|*U zLEGu*bQ6JGJ>%fZt(CBJ966v+m5c>t`LPFzVi} zxB)TY=Nu4+8Zz30XRgjzpVbyPDb^OweqsFleSFlBIYEROmb`}rar6BLwRuzUJZVaz z8&3n19*zoZP7FwdT%7TUKCcYa{lf|Yg{Y6}=(igSAs&tvZ z6C`>XsJ>aZP&XZ@SPx>~;Y7imq&Ap-vHB%B{UoD%G2>mbC$Sd-{X0Y1E7GuR(9KH$12;pV7(d@Dp-G!`FKd;>& zNtJ4420Jyu4K3OUZX{bC6Y_n#HY19Dyj766dbYx*gW;UIYgxXfM%bT&aumR_JT#8R zp`rOMgY{wUHcASXEsl_bEVggrvBQt&(|mkqZz$uOqugq@vL z>N~Qt-w;k~9P41Ao)cU7j!e1MBk*XWgV2`ly4b%L)MVZI5ADWEs$0lz_9kG+vKk7d}AA~-Z)wi$ua|SIxRYqb?&Ja zx*ARetk;oJa|p9j>u9TzZDK1;6t^p+#mHWNQ(AMV{Tc6Ww+YuCV_D@~T!?BnWKpqB zh$}qQ+dCEy$EI#<)@EEme}gHB(2A@TBxc>w9E?$uXj+a~**TDNMBk0-h3?C&2nVzq zBWV{Oj8TMWo{k9AIypy}f=I8Ao7lZbn^Oh(drV2Bj{E>H#r{v&2j4E9cHNoc+xLTt7VG5iR~5w}K5ojqo*nAK)ef&3+=B)TCt z81F-ufd?D|5+N5e%nwXA6AmBqPU@Bv&)Z)p{MB|wbmCwTU_ZjQ=!k{GUP*`=@3AU;NV^G=szQH^DtN2YVns17CvkDyF7{v7 zW>Eq9S4~O${{nQEtK#7v9j)Sk?(rT2{ny$}a3w%@IVc|Pa*tl5CKLUYHj4_-e`!i0 zb)dIBc1(2TyAg8cr8A|AM-fi)0z0%Yrx!Sxb zn9eaJkvgU;Lo;k0qT>f6o9~oTM(b8>aguh#Rk&}}&-3m&4sr4D5*3?1ZB7(yZZRd% zjg5g%7rKSK9&11%Mq;wSL_dq44;JEGZ4~2Gk{Tg)A^soY7UD;>zpAL;h#+wl15kTU zP_k^UmGzgZ4pk>`P#Yy$ijcw~x_qs~+j0F^OxIe0840!>KcwAc34v`rNnEJ<<=X5j2!FpRiLh%wBS_48kZFyKHBO*XFNuw*z%c%Twopl7 zT#L1a5GR}^MO+1O;3aYKp#P@Li~`dCV@e`1NXE-{8_!oKoqA3jsb|3kQihUx9$(K| zjMpD)%a;_dO_?V1C%RlMi(`L0{v^<1B03+{=12kUe+d#-(RN;c9)na6e6vPfTD*+M zWBRHL%2%|TASoy}h;4M?&Pe7o7MVD!_b=LvDcJm@DT%N{{R6{0+o|!J9>s=b8vOhm z4(e0tWj4L5F?j2@&d!uY@i=f)`y_K{`}r2stD@K1EI}GY%+|*>+Ds|Htu!T(I&i~> z3(hzla^#M+ktJf-wvo5w1aG<;{xw@Xv?8!)Lv5(-5O;u zJql-7JXjcz2)Rhf4O`DqFFN^?OcBNyb$IVlhjs0ub^|BXwG9~#gMl}@$Iu>=tdrN~ zN6{JAltk)i9XjS#vBj*8WjM^FaOX$e^)7#;wR;?UeUbO!QSy1KaH-Vxei5aBzD?o^r)3 zUAi;RMtAUs zFe&W?7JFo6&|)^WeNvkv1+-5H5?67`wemzIOWSm?vrP<-Q3#Y$Zu)O(3!1d)S0Tt0 z@Iymlq{-vvQs2<#MQQY}o0157%2ye;*`7;zS;KJ1bL6{3!Ran!Y_@JoYu@Ci6caZp zG%>TK8?{+cpxI(dB6Vn%@QoZQEOG%ua#UN0q}^qycjtx*M@&TMkTx3%9QT`&=!U~! zo)Ato*hm_X2)RgUs1Okl!c(p8t-@+CE6AjFvnExL)!tEm5=WAc7_-9O z)MiCdlBY~bbi-qSaHu4_@MAzCBk^L_^1dFomiI30uPTc5PC?=-I!D(ZburW-2u%wFXxHA&+n&BEck_>F zH$lRNZ{Ls^lQ^}RQ~dB0A?WhU;ykE7(q>CR<|C#g!p`;w4DM`uP}d)-Rwk;Y7KH`S zH@q%#np-C7PfcHyA^e)QpvfVmt6Kt2yD$vh;AL@f>3`K`OM&pqrX*5_aPwI=+rpA1 zk`<=yg45Jv@UF|uj%%zKzV>y!nI#_gI986i=Z#1q^D=C4 zzR4@JIaGjqnJI~|t6gU7XWKVfcOYnpUYF0sS7G$tsV!Ji^wwuW{fbNFT82$TG4Y;v zXfvdM^>d~qy1_D-_=R&v?&}65LM~=-XN|2hEnixp4YVW^ZS|cxc_x~{*`B#N>(Xbm z+dZi+-6&FJLXIV?ZI??m*|CjT2ll);8w;b&c>Y9W=%n8*px(QV~#V(v+Z5( zJBWNVabQ*fr!H2b?h-#`nrpgBGl9wLmCYHgm%*;1%6*<)RWCdK7}tE z4#9wE*!qsOd$}_QqD9mp#(!;SGpwjX&6Gsy>TuK10tG;$z#~gHXq`w*X2#kK@f)?p zO$xD<86$qX(8Qe=0lyJ6W0pJLpv{|t=JTc`Qb)62x(SL*^ug}2JIGCs4$m6zF~Yy4 z-2_Py-kE8HcR`WiRyuAdLj>!Xh}>^zbE#nc0aFs)SQ~hHp|i*Xz5$7l3!ZrzkO;Zp zg*F2cAs6p3T}~h_uv_xubLY(I>G#a~oNXgkiJG*)06*t}7Oqu6>=S#jFGe~~@xSJQ zr-$gD1-HzOXG&6bEungfSw%DOx8`bpOVRB)rX)f+#w5ObGzJS^JNlL?nhcsXv{l0Vdx%0z&cJ13Yv={$a4j)tfFwfNs%_%x* z^ub1{I$)iq#o(!^UMvf++NIsl@z#=BYbL!I)c%&D7q*| zN{YvtjL3>ccf_VFk0%9`r?fw>fKo9fkvb^2&%>?5xh-G|gE#jkhiJTBZB=p+d()o2 zDns=KZ9$Vlbpx*ckT@0dZi!O`wdb`tQBZrGDT&0W84olfz@EiM!iH<*anb@ZZY(;z zixRPLI&Jqcz%y58_&%g9aZ>oU@(qxJV;daX;Mn#!wwYrE!e7_sMuG5G1&OOT)Y#`} z#SzEi(Y`MSQ%P~i1vvm!9RF7?JT1Z#HEe22~-L84fy#SGUq+h@mxS6sFBYOH+= z7n67TU`*>Mi2Rpna~9|O^Me00CDDzD!8|XVDssOzAQ5sggO_7$n%AGD;S?wRTIVSr zyIQJ_7rI@Y9(7nncBQ5w8;~!?yxAf%+viQulR<43l%9L5AaNBf?`rP~)}Snp{o40;&J?VqEkGJAN2G*|M>|2c#MC%n15Ul zKiUiEUWH1tQxt>SLULzGRi5bdoo+8`Wb5U1<2?d+49{;vcuYsO((IA;B=;dEndlue~r!XES31cm8+U zOX|*~^yj?sQgyWRTzf&jkjuH1&a>_19(nJj&_Hs4wdQb zHqYHW>N?F9?xBiI1d7*ze|9V#C`D`3(eY6hiBnQAg2EG6P={|@FLXpxbgi3+rQ_mO zYO3N(;NBF&m9;0ezp8kWj|&o4adH)1CW+*1M2om#U#2Z$yvI&D7HQb-6_j9w{@S=i ztIOJKC@uZ8DT%P7O)^-sJ+-n=;NYQp6PH6{ZIL!Th{QQ>+Xn9M3sOsN$;7*AyR zdm_*>8PfW@4!*uN{t~=B+Ds|{-eyW7>~*#&iEy66YIH^O6n38sNQ7LlyKF!rY>UTC zjhHuE`pc_3Tk1Pc0aBVB`EgS|z7i9ri^vwB9byz(z%)ko<%|8n)aa$!*vEGhk~`wT z=eK#P8cv9&w7;cj`-C8&Z?23-j+KwP?#0rSbw7u7h z_O|!zvmL>i&73K!{p;EsD2@KBrX)h+^~;RgY|lV%V03Qx(b;pGeGaQCF0F=Gd!jJS zP?z!gytb%G@w(B&>vkV6@Y?WswfCsK<+IwXD8PN%ltk*lZD8Q;utKiSOaVkU)ZqX*R;5rf zp`<8n@=&@{LW%iPV7f({2?eHYrX*7DN*(}rNiglc&ECB?+LwCPVwg^9i>C|IT@p;p zp90efZ6*|$j+&B49j2Q^Z`ymC4AUL9^QEh#)3Ja~2o@Gx%*K=0FLK>_H! zrX;QwKzl-f?zDIBjs>W@4g>TVZP9cA+7kjKcvArSlr{?rK!0vZ;%Wi3Hvnk&K703F zu>f_~VSwiBOv@Gbd(hqhAm&X0=v$_(8|Ib&hbf8F0pbY5-hJ4FIl*(=U3Qm+i?71) z+?*ObUJzpMKAdyMoY};gg3eZL4wUY5qbZ5h(eVbC+e38jw0FhZx9&QO&v9+hbnV;i zAwGgP1)w9^EGPgSG$oNbKpaQddq)V+u6_2dE(6!A5<^td7A`47Ua(>B9U(-_oq|z8 zn+FA>F;fy(3!^(jjCSv~ci$czgglItN{rEKwT08g=*|!$=1#%rHQGEV7`@t*MCurM zlhV5+jP9~&&6LNaE|WK7ZAR@^w8c$|nm>iTOG51~S{@bb+Pl?Z>X)>6QP6vrDT!|M z4Au(6Hgw;g6h#BX7@*9MTc7b;Ub^h0t$kt5kdsp|oy;*mT?9ir{UrxG_U!U-ConwW23a`*4dfRC{2z7^K;CjNPvp+7^Ku(6PHT5 zUYjQckTs?xLbI?^keK!4o7Ly5$bv$$fH-$CSIBmqxENE;NbT1aDPiZpZssQQBywm3 zOAK3zo##Z^=#Cf)THv>8`a<-eJdNL^K~ zgY#--TR6!c$;lIR?Y-|q&)%&&iQP8^Eltk)iZ8~zPhRZ+T58;HhNTWmm>s^g8bF~FbirMCj z8*fo;T;MC!1u0c^NZjE}#Ya7xN8#c>rFs#j|Zl@zMA1eMPf=C6V) zsc$V+L3))oGYUvQWJ)4+knYf#my})Xn~Ouzm|M9(`O}0!5Gh*px)-dbABwA%99}9hUrT>u9Tj3kj9GK9jt_$bUmy=%mPJ zXsYCM9U02w?kBuvc!dh;*R`2bK>eyIiSK4mZ?w~z9L>uVa=YaD&Po-vE!tmKP}^im zB6ZaEP)Mnn$Dw`Jm{X>hqHfUwa`%&_w={0gE4(@wUoh!X2XUb~r+>cu??jJO!4SMqA7{gRptI9s!~Oaxo!m~dQB6fyK-`@ET7bFu%yaz ze-dR80^5PWJ+)<>sx6<;=3LR1kDHQ6U0e1i(H1C+HCm`t%hQK*uaFk&$v3r|Dyg3A zPogKZ_Tr1$Q%TmVO7abDwiPA$x+#h83?(^JEi`bZ_7`6EQw?hxdy>cv^=R$rro<|; zbug`I=iNzEq{f2Ujh_0kLDi2NwRu+bV~Z(?7n*(?)o!SyCWQRZJ+dFVQAPTY_SY4p z?>8lpI@0S;xh~&;hzvPDWV$sPHSSS|`|d?;(UM}bE>i|?Vv-oLCS$cp#VW7Oje?bH zN+NZvcEn*t*O>AFcVwf-ai?5Y%s=s4jOS~$#ZQXojyODRq$@J}WPF)@DPX@@h4?kv z3@Q-6+LT1<5HC3tbi!yKlnWTB_h}1}6sV<{`2KdC3dVc2nNYxZw<(F#!RU|cZvz%q zPO9~~i_}0|H|0*ynp+erjrSPJPir?pQYiak`P)kTMhM1+k7Aq>oFTA(HHp{53}T+u99~6y<(8V3;MaB?*G0a7h9Q=Id3M zzp2fl0`oUaNqn!seDhFRlNoYW_sR~?4Jv|LwZE<)c%vza)DgTVX-6HfoNTj&{)`KD zTO#|Z%S2xhCGO5Aw3{hucfKcaceYtfTM|?1;5Vx3aa5aWMLiCgl1N=W?nUoyZla{BaaVHHfQYs%qNEr%sfuw?n`K2Y@}?wGSB(A16@wz9 z4X0cyxxD93NujzzTC530XMdeL|B&wt&mZ~JL)n;2!lGm7$NL@+p zORgk-#vR=aAc7~j{-_+C16(mRR+RT^H&;?cxi7h*_}O(rT#CmCaV1sRs;bKSv>8`a z<-MjPQdgC`596N8N`dm{Xot}uZ#qbS26=U2`hbHEDmkZ~!-*jig_wITu7nUP$Y-@1 zDXD_olcC!@SxVZmt|MtmtH<=Cbw$EUTKR+QtDrX*fS%5m*& zX--A<$)iO^MfuyN5QXQ{ebbc03kl^NsiJ(ljPlJY$~S9&UFpkPO-a0vP(G>MNJ$Mn zg%$3QQNBe*`Goe@6_k&fl1LrpVe{d9VEW0ztWxFy|0(TeOsWUNS1{mDH<8$`s#8T< z2t}PPnvzIeovy^Gi^L(Wgb*vsPiQw%Qe`2dai?rFcBm-7Ui<3`%C9vgkvhtE9;)V$ z&|k%}dAZiYNiq~U2v_g?HCcGZT_w?ZqB$8x_=DOFloa8+unL%S?Pe98zmTwNCEj1Q zrAtF8ln}qj(5I@!`?XnC)Z%@nBvMz46=%^N)v|?P0xk{ciaK`}GK8Pk7A7f#D>HS$ zja~7$k)~z+DlnhbW<&wz)21X+2WEE^m;r0lZQ?dItKyX1yRBNoZRM&KH<#r%SHv2d z5MY4k?@DVLwyO(Zn}xDTD4S)nd#k`sRRylqW>ryvZ-2+Cz!pyh7+!1Kt+;ikS*=@V zPalg~x}e++6|Hv>8|b@MVe2`^dz!>T?(U8*P!3VzzZ!i(Gdqx5@NJ)Q#w%uAMZuZnL|n?D8Lr%Xwtj_(k5Q)4pbpCx?GE#croZ2RzF=P0jr z%#S2>O+~cdX2jp5-5f~~zm2xTdh$?CPs5WZ8YB$k5RoH__o>+bls2ad_CH}t;(HVO z-_~x9q}YoQHnP|Cb`{Xy)c(8z=m$+nd~X8!_u9>&1!!bH$Q>%6f1~|*1<=o%l1LrU z{&P701DV=6w~Uhvn+WE68JP&^H>{SpU8gqH%vUSXfS|^EY>}7lNo%sQGt&?KsOy|;SM7yaeB-VgQituZoa^pGCf432^M!0ax_}<8ur{nlzS=6~tP`-xU9+LtA<63Sw06TKRfi+U zhT%=$a{jwmuo05?^q_5)+%CYDN2XHusH!um&Ay^KO;Zx7tIozpg@#~gjwr)@(GK6U z76be<+TtYz7?%Cj&gzR1wvNvfnq(JslrX*5_Z0qsrB&E1Pb43!{DP%Ocm5Lj7 z>?n8%>NAYLr!8_)7`J8e)k#WP1J5EY4g2B+&*7w2w%_|zM1NbGHwDq(G$oNbqRSu6 zPg$d-YW5;-p6u#+U1f~am$XGmiqvv+xGCG`MF!5tNVyF5h8cZi}^aaJgD zZ``BZY)O0L;UL>rlAH5-L%C^MN~ykGa>shc_p5qyn>P1~-q@xjQrDZ+kLI1GRcQ3G zt}*3|>0{a=CB<|#EUeSC3yoMSdq4%~VQp3vfXcbZ<~@x9hkube5z5jMXbp}GcT5z2e9i1$3e!sA-XR! zu>Y*x7)gO0%y@?vMWwMx8nN!ocUcF*NVp$SasQGwp9=1OZ%QI{+}E9lvw5&uZPFS? z%__TC(#S=f%=jvd;p%;9O&8XS+h0a~&TNk}396$iR4cUEQJ`9CN+NZrmIpA^FvE^w z)K$jd+^#K3(hjyF696N4kx)6NqOwPu4+WLmOi84U%K8W@16JM57HVB_=pJ<#rN^~J zON!F^C`z`+ngrHy6|Bd!*-^lH*px(yVD%M;=gpbZUfG3Kv==F*jL^%3JWODonsddE z&*%XKHl$57UNu&?u z(g&RE#f!N5+8wZ>Fh)ukm;as|Lav$ygex;epJgTc?Ew|OpEqs$5WcsYlIVs{r#U=( zFfxD69B%xt?j$F{fJDf}R|E-1qhCo}v_^}cnG0rK0*?A$-l2YZ41F)Zt$3V!qT8dO zuXwRXNwL{qk7d$#sv|#iJm=&GRK~km79y$=F`vRI>SMop$MV z^4q*d2yYVqd+l#2*8Oh;iJpzwSM1fArj5Db1SR0Kvb3GHP@$XdW9E*%DsdApx_x%s zRHcdA8#9G~I~+J4a;G%?d~F_-roYydMA(tO&FIZGM%#tLL&>cJjRr1*b#QXIb@c!C zb}m3tR96^gVVQlgySxQKM3$$pEH1AVML<;41w@5t8E~APX=bLGof)U6_mMIcU%;$s z#gwB)Nn&9V-!W*Sg(8e8I30u$LYg8N$YUh5j%gF-Nczxmt}W__4T1{>+#9S zL;_E-s`d3^eUCbwa>O?EsMCqPRdLkowCqrE%jw>$wX))f%|F~)DX)vVakEFUFFWE) zdldV*y;bqWV(-PU=8udTOM9Whx8lv-Ms!+%k@5BaHZp$s!MtLn(UHZ)THjMjJvhR`P5- zvVd8MewSxeGTw$sN<7=ica5HBB~rx!x>p)BF5G{_#d4Un=*+51&@VoS;rWHxwD?tz zTT;Ac>qz83*Nd#~-(;!(#S-Z~P*l6od?STehP z6HUK4V7|;`r1=t)G3H(-W#$V^%FQ>JG?_0lsWe|#N5ZE z(%j9Y%zTZ>AoEow%`&@gE|pNy+}Cz$Xp?1xy?5PU!tGRIPLRdKARZPIW8<1j*5}PB zQ*nMtbIFFhxm>X?&{FIN{I+5#@Cn8Kzypc_U>Uxp4h?0%X^Q2*^A#(AZHkq^&5BjP z?TQ0{Zz)y-EAdr5D60WBDAod3Dh>p8Dh>jEPjN8tDa9eccNB*LtMRo8C>sWBR2&Yx zNbwlp6^bK(TNFnEcPJhUJg7JdI1q0ug|gAW*^0*jFHsx={D$K3z#k}%1wO5K0`O0Y zf0+FDuRm{zLIJ;PF_&hq490GZYsBMI@SQAKv3*A&@-`b?Gs*nuV}vI8w~vAJX;)Ru>1p){uGC%KCILGi5toYzkjKdwl{ z`xU7;P$maZ@l-`BUZKdIX)Cg4-mb`=`Ey0~%>6FX2s2PF>MEN1LYS)p`|N5(_Svfx z*=L_nWS@P*MbR>_LbR|0FL6=!Cp>}LzpDbf_LGY2+J8`F*B($QTG+LxyI5RMjc!4e zqM#~kad4>=RD&W5YNa9zs?9|)z$`F^74I ztmB{}>-btl*73cHtjT8;S(6_tvL;6i5oN5&`HHN(7DZOkMnzW91B$Go=M`B+e^q4k z9W_*xG5Ri4Wb`!@8GWx)Wc2;9BBSq1ij2OWDl+}8GVC_jK0?@GWy=D$msj5BBSreij2M^Mv5{< z-}#D&zJ80F?Y6iww76UNdzrMjZ7xc;c-<4&KWmPagV+{lDzYtJtjM{Xp1NCpdFsj55hrDD1|WkI8jVbPIFOC4|@Vn&#S=x_Na?I$S>@3T zIA*M~35vnSA|D^#F& zTNLTg4nrZ}DLe^8OeS93V6;xG$+r>OuF{h~P8)iway_A|c$qAgG-NNrOUscnTKeY6$nU{Tn@jCqLi`EQ;qn0S(!9D!l2$4A=TB0IKPRI-#7D6(ls6xp=96gl)gsK}x3MMVyM ze^=ztcU+Aqp~-e$CBJZhp_r+uXdv&AZ&Z$4%)ZYy#Yj;bts1P29}G#;}}$ zYvT#S3T9KmjL~Ap6A@OrF@O(?q>TGREB{0mOu$;h!D&3TKZ&jUN+NaNQXMh!N!R$Y}s3p~A<{3IL!gVuTZ+jNT$y?vzDRVYwN!+}w(p@^Ys(Ze3s$9@FJi1hM#?h-8dp zeofw~4u+Dcp1f|u@Qqy%lXk9PB9)Ub@KxKf5*PqFZj@^Cz>hOkZO zi3ktwP9Ex1dlH(@TJEu(e&KYQD$V*PG_$u*DCTrfae*@P`_*3Xj!OQXD9U*K?m)SuBI zmbtE!eR0tRXLHd72^U?EaM1+`7hRBW(FF+?U663m1ql~jkZ{oj2^U?EaM1+`yAKjB zx**}A3lc86AmP0zBwXfP!%)!IWbsWlPlfp?*s|7C#757=KCjBmDY(WdIHWojvVz9t zwp5Xx%$1{@_>dD9a^gWw9LS0NIWa#c*5}0doYG*r4P{Kg(4X(ud*etXgsKJ|AIy&<=p)jw+(N1kJ5&BwZ z5D&=70S11F4Ry-&<++Kx)BhZ47}@+)m^f4YP;)&Ja|1RT^BeLl8}m*%szWV+y`A;l z5^h3Yjl(7vp6Co%CUrOVKj^)oKySb@LYZ(3Q$y@Lq_V0D<%qE>LQx~%!^WEnbQKw9 zZWe{kkQMTlgRw;>ahS)Q=2X%zV<}j-bU9$*Z1Z|i0g1-B7DA$CS3L%??g}l9CTxRE z*y$hU-E*K5K=^76X*gip+@u{02|E}Pb}%IDU`W`(kg$UxVFyFP4u*ss3<*0J5_Tyh z>`R+m8^A=?Q+-CNJ();_B0R&=!<@mXOgzdZoS>)=+L<_=tYCZ!x?_V=g+Le#r;=IR zDB-VFt({huJ21?{0+G@oEaPXUb4bZMHOnFhb&wb9F8! z#$#)+GZ-=C>vp)&g83@;Vmuw5Y=M0;J&OR0a4?M`lxf5W{3JDkiG=)O*=yIvJ76DX zAFvLk4K^gYNHDi5GpmoVu9D3CY$Uw(y3ar`1*8y$)Y`HZyoWk zr&l{=89S-F7FMpGf90s1PdNeiDce|!Q(@WRuz{IEB%iZoL|}-17j0WKj6Ct@GdV9yQAp33&C_R<(}YS9C^~7VXjTVwUR#V|$TJ z6Zbd2=$357XHWj}z^?6uUK~}MX1g|VpeeFx(gTBk>)KSlaQmfOo?7vLvuTcN6Gy!w znB$L8rbgZnRG|ExZHE8D4m(8EOkz;Y>W8mB~#zCOSn{~lM}5`g}-o;7r_!r!S939p(HLSf?*5)y^?Z9 g^I?`Co;qkXa0zSSL4+`tL%iR+Vt2*vvs@~ diff --git a/doc/LectureNotes/_build/html/_images/chapter1_11_0.png b/doc/LectureNotes/_build/html/_images/chapter1_11_0.png new file mode 100644 index 0000000000000000000000000000000000000000..9d454e64faddd98f990873cce3da89a3f8d635a9 GIT binary patch literal 6285 zcmdT}cUTi?w@-*9D|P)8g3>Gl>`D_sBHe~K(o`HEbWlJL2ps|fK~~&dRB%{9ieLf6 z(4urgiE;rW1dt9Ah!8+(klu5n`<46M=lkxz_ugk_o_XfXnfJWsJ>~tKGrwIjGZGd! zAb>z1gux5vt{@P|INl}x1DZ!kmY|CBZd!qc)`1oTx4;l*e^-R5bD*ymA<)ai^{J80Z_It*Yw%cLo)Lzq{)DhPI;7Z8eQbK4OY3$Np^Q}_xPXP-rD z2aP5n&m4Fy&S(`4a|VQcsmm?MX$b_Icpt(5lP6WK|?Fu`)r(R^VDXlnTq z)M_VDME*pcdQ9HB3_m>LJ?}oHOw^gzw_oYD7=&Uh^&HUrYnbfLDMK zu#N>+gp@d9VDO3f_=^jhppWsweSl-K(0-Y7gkB_HsYVZhXUyJIaw;z_OoO#Tx3uV! z;EEj46fQz>Ul40TfOQ-7>Lc{aj4e^mx?c->oWE(5#NNg7aw4cl>dxU=7bY$0|pXrwnP{8n(xv<(_DwfzXx-ToBQjA!^sJyQ9Z?y>w>gW%OvY z@%iez0nx*JJNqfS*A!XXC`N{`E`MLoJ#NhSwl`V@;egRCCDu41%=jIKaa^Zx4{d0S zMR|^9iL=6KA<=yLBD&SYxf8Ta^Z*C#FsQo&%)i;w9HBYiAKj}}Ot@vd7-pm~suOLT z+GLv+*3&8YjEm9DAq;hriXMbAF$#0roeoD?N@ic z@_U{f*0EVu(j7~(yH)q=CCgvpQJR{aI8@us;M2@UY!#L)GX8+wDcDH9@+e&z+{qaF z5H;#VRV&=pN{}&-UF#%Q-5u(A8^)@tJ+PCXqTk}iGFUF%)~umv)fO>kQHwK0WZ9+r z1_vri30pmG1Gy_5S9$Jm16od|g+~t)NRx(nU^2~WOc&O z#f_#o0&(2{qbWx`T+?pd#~-9HAHkZ~@C>zH7I`lkpfi6Hb;6*8yc1@k95d?p0;?@( z=9y!<;gI#3Cj-qvWISYJzj#Y;U4-4`@oDTs4CK(>*p?8b8YVsj*QD*BH@}4=Wgb|5 z4h5;sA7XhgJ-}=c)wd&AOq#4bao5s=Y=Aiat6|OFP;WT%=77EueNmn`8ZjhVtxwMZ zBqyb1`?bc7XFcRe3x!H$*N%iWiTYr~MI<;cSFJYl_V3`i@wCKf>ZJ5OZesLARPvC( zS|L!UHZoSUt96(DRcLkDG)lN8X0BI*TA`2LYWa18i=7u|k#lCTU!+Si8zrOlXv?&T z=sJ%a(D*Rq^W-Aj7|_g%PI4mUg1;r42G5xKe@&Qhgs(@zZ71rV&oPgHsrOSHSB3`z zXOFY?)AU&b0F^uyhcY#|1YsFfGGND2Ro|PFVtuv{&Z-;9IvR#KCd0BZFMJ4zw^E3zTEogQl zXWJTz$RkJg-P^&j{t%gh0n^+#N`-h|Dw_>VlNL$#+rseDaQTE3tMU8W8{^bx>c3MBj-b)r>pga|7#>j!TBLOAHd1=r?E%WW-}7AU@y*doQ9 zzGS33>Jw=((Mz&p&YCqwb9b{(1U$6uz3}M*GX>D)1}wWD zkOv%~PdM)Zjm;4NJcV=i!}z)kukB`(@Hg?rWAszl;;dIR6;XfVj;$|;l?wUGFQowl;42rqB90=03coJ%74+~x{j zcJqAH{x$#avNaa3i9pc`*EXqg zo{NATBP|T157@EwcL0(1>fK>_b7D1>xFzdm0RtXGCREq25+V-wBWC{6dl*c$=!`W) zikdEjvOcQk8o?TObA;mz5Pa1?tQBOd@}A>!YFbB>X-(0hP&f0h9LzVN`tuNr0jgx4 z)`PROG`y{KrTPS7KR7Mb0)Y&F#uVz+=utinnS+IPHslO^r;b_IzIPQ4?6h;M?q7 zmm*h|gqKuW69H*7Pdx%c^jW#&)I(#(2Y`o+Y53(KF&dJEi;h>0(1Npw8jSu++t|;u zjnhiNBt?xkOL-CWxnqAm7!s2^skb*}aq+me{G(up-B)+sEDN^q5D&jaSwe7^XKL{? zK3ytVCBymq+AwDIkJ44;wPY)(DZ$GvM-u_YeGk{QXt;vrzZTX!EvQ zaj+wO>Uzi(xELP%<2nJx*motfAAz-x*&BeY`G~22oX3L}T_rf5Uj@~QE<;Wz^lP@y z6w_A78$?!obb0HaKKM^t8f$O|4K4phu6nt@27U5b3@_65P90WyFv0R>AXiRwo+ti1 z|IyhTs60r|0K!)|!*##N@pjgfJPW79>L2?=s=fObq0$%RGOO+t(=^Q>+jKy>F{cLC zAR~%*0RYP&flY@t0AeOl-2_%DBv;an3*X;0vn`JNde@kH0dq? zOnui<|C~IBQsE0m;rL|p$&#>LDmy)={}bsNNmOp#4z@oct^nGg?Pp?b)|hd5m>m0# z)wLFtsRTf|)lQm~1TYl-B`p6JNd13<^xptUW)hnK4<@rK?Y}8M*uwKH8`g0is+RPP zguPJ=ACv>1oz`?1FAM2BkLmprj-Y@_fs@#^kux#ms%+xij>yzQ!1Prd^HLM4eDyx( zW{#;cogZ8Q`irBuJ90mp!CgLv(yT(VG*ntcE+Y(-cvLn&%&I^Ck7TC-b_jelw6+uU zGWGX|gJ?urp@@SOTyl=LpF%;hzI!N=6?@j4AZ7x)1DHSn>kYsrX#&Z)=wF{d&CB0~ zamCMa$$8O#HHS?Y^(8MJZt$cg0m`EPlRa#h&VCJI6j#A~eCOus?}|y(!=TMW?aH4B zcsoCSJ745rfH@Bl+lG%KR72k`bn_rs`xGyqF{ee)4Y$MRCQXw8Q`hC#2JdoHoIR}3 zF(n+AXqM_z1$bn2ONtZ1_U1csX&%$Pt2xGS&YQ6M5q?#Yw2srF+l4UhT;J!n-k{Ad z-Qk-1mTtfGO~^W17FN!y<%}EuQPst%5I6ULid|FsG(Gd`?O(IRT9o4oL%-1;JAV`- z7##NCS68YUnKc3moMH{F{$!y@xQV+ffB*8a`&r@CF}(3m=Z=A=b>BM)F^v~x=g*f8Q=Uz|KO%c__WcnB$Is@S2C%lXbznAEb?kD~ zm}udb-XcyxsC%XB$GY3H%Qi0@!n-bRe|ce4*9dHDm~u__O4t($Og6$4T@}gA+Ey1c zrOpe*sR-G=GSl|zJGQPC_eAGzQLd46vGq)Q9i%>EtmiY7J@;~lWt02DU`_nCdtZ-P zlz~d{n1*MqGspGNjJW4OHSx_gu(NuuWVTVypzq~$kFxy}oe=%`_(WmJ##LrT+ZM3g zcGOqo#U-AZj?w3y_$iJHs_ItU##761`Zp$8F-1by*gGp&hm*{k@f(6s^|4y2i43wp0@m@cwbW!@mR&v)x#<;hhaWIZ`+<;?V*2UwjaP851F{7h={?9X@3gk>nP1j{v)J2Jt~H zYNA_G>3)r<@lQYCtQa2F2zMk786F*XS2t?keDwHw8^U%VsZt43A?lj#xAZZw;C*93 z`uTz%751Yiv=-j4sJ_o-;=MVpYE8D{w{O1Qrp`y5pnvda&t2^=YbK=DavXnCZaBU< z-1`hYz?!(((biYNop46>8aQ%8ywx(3?;s>rv~)-BeO^@B)tJR8;s&T!l-9X1<`LR| zsQ=u=7%rQ6t6Z&CpZM7b3Xl-z!5l@JIdYm*$6qE~vR+QpNo$%w%HFV@dy}rML%*A6 z(5xhhSUZTBVCay~)wE0nehEq@$)q@151-j~Lt< z%}84ieQ)uUJg%-}h6-N2Und?qdn%1d^I(h~A?ZZlvz=>s=Q0@e0dORF7|>$e&z5@Q zOP-8tPdaKQW;{9JrI_bE)Zn#!Pw-ut^Jl#WIt8HvT+Z78af@C`tyb5V&}NHrcakBt z9&rE6@P9n6qYZLbqRMjgM*UJb?23t;nWlzVL$ar3?TP~#7>kK zzQaE>J=tZ;H%YL%RQ^^uK6do3a+I@{h<{WT9AnvK@1Ez0p`dukfQO%3J_W+X)g7^`u~n#(uL)bf zxDB*xf)5Zf=v#R)13ZlJUrM1c3?)Rkq@TaKg1BYJQuXZFY~nGbt^4GXt5oTROCl~s z{6iKH4ygQNsNoQm#1*Mlr8|Ues<)*%ebjNYT>5a$QR@2N&XU~P$Lrq!^@j-#{%>V# zfUG#!dePqkt^rkK7Pc;f`t7>h6u>s=S&NxTJGRJVp#kVsSPz96&6nM=0~^nam;vxr z7$Od^f$ao#CN49MheG;;?%SuZHvO)^=D@nwqBOC`zpMWazyTK+;|I!|+Z5JDYHtik3*Hi_K=;)S2B~(syC}}%Sqf=i zjg&%|&3#xMl3-fXCz5Rvb#DJkQzuCQA!y z?aSpSG@8!0#+IV4hjJ;gW$W7Pgh zEJ(XXnPDzRJOeRH(+!mq7o%4#yJZ2b`$MJn@kQj#WGdy6(@?omHT3>;;u1gDy2}BR zZUGt0MY>MwC0RTO zRjQ>)UmZ1{^oA0yAE{wp!G%~~v*PLciKLEzx0GrqeZwA)7O!FFRnv-}X9x`1Ls2-N z*vcVcsF(`9|6SeLv0C^fdhJrcJ#HCU+r9tyjEB27UAogKvcV{>1|1tMi!3T6L*-+7 zJWheo5ZJ0+r*9jEHtx!SL*|_LU+y#1QgH|)zN_6jlhVaR@ft^k8qAtiL9<^}xC+hNr{)IACn9=*W zS+OEp?hm%=B-4LviM%oA-+M$^EaWU`4AA$IjF1>1fdXEOx@1?X<9_PxA=QKkf*ALG zf`~=zoxoq=e9X?SPzYjP;7r&JQ>r1*xcS<*7L(84L-gPDF1ooG(U^v|%aW@|-md`! NXlQn>@Ql;#{{&!XAQJ!p literal 0 HcmV?d00001 diff --git a/doc/LectureNotes/_build/html/_images/chapter1_13_1.png b/doc/LectureNotes/_build/html/_images/chapter1_13_1.png new file mode 100644 index 0000000000000000000000000000000000000000..d0c5b8666a9be261ac4521ae707fbfa0fdaf0204 GIT binary patch literal 9553 zcmaKS2T)U8^Y=|4^cJatUR4md*?s%P3BI{y=Tv!b9Q$3{C4+5>+5PT(s9!P0KllFscHxS zpm6fTK|@9U;=-6PN&b`dQ8V)~a!2|2+j%(vI(9xDuI@grPXECC9K5`p+})6(^`_KM zzPWPmhis*bX>eYTR^x8t3>HNNg%*?;sBtQ(s=mxsg24?v=_uiF_Pqby#Bt*;!)rhM z^7;eD6>4UrwRp~GZHM6QXo?CP&S+V`*$c?A%=e`17s9!KENxMWVs*j>n${p&T8uQX z3M%6!{DVa2w|lW;Nee(6WK+oHpjXdJ3Nu5&<$^-fUCUrol>`gPF+DnQX6P z*a5e0;v~ZS!?{K<@IaahR8S>;3W{0M%((BRGPUob6%tKKLfG&J(09=KZiL2?<`LCN zC+r>jLyi7I9gY!ftydy9nvTNzH8@ zga-8nmZ|@$SmKF%M~s$*2r=PA5T~uBKx4~^`KrmMwQY@?&@AB(@Usf&WeADnZK-Wy zbIrWB9Yj|$@m3Zfy<181xl5bj0l*au=w8^r-X#vjZz@ex3@fewT*#eg-F)C>(#4R}G3 zLP=x-dOL@IvIi!$w}p@o8Hs^_W9k?MVg^LC zYPrQhF$&mPY@X5rUor;YP;PY7PMW{9}g;bNMUeCNRoKyoNhY9qUBx zt<8w_rw zCrm_?9s|Nm{qTthp1pSf6BU5VEY{|ELhT`8S<<{%%e8cti07p-b$Lohn5^@Thm;kp zAB=!g_!>ezzHWW(F=;MlCTuY(YAa4A-3Ef_L^q2B4m;~!b(#PfAP?q_nZyQ(xg{Ac zLS5a)2xP*!N=mL}#KtkP>J>LGTaMbc059?+DB{s&%mzq?ouGoMdPABA%&H%*w(26$ z(WJY83gTQrW}gT6pq`^71(L~GKUVq5G_hHJ(%+>UBXceQC@6agV`4xjG=z7kDhg5g ziWE{G5Td*dMb?+jvh;q1q?)wsmxn)bGNSNOD^Q1NVXw%(TOV3h`MwWWYHlB7Ep;Ps zq3T|WHEH#rQ81u9ztlIyHrzD~)bJ4AMEFIb0bzk!#SR0+Id6v|t=~Bm)0>p0wXFn% zjxBEb15cSdytU#Otsw^IXJSVb7-O?+{F#da>uu|V8xzZBs}pr3?x7b(108*@m5WyFY_e!_NwC;i0D%z z6}dFqZ#M(JsW7@=3AkuA(Gl;~R6v0S9zyS|*AX+QX1wdcv$Q!L560hm-~&^jnW-xG zN0%SSjE>j>`|f*pe2GQTNeQpXgHWU+awAu6)Ct4Q^|4AhppI%t@OfK88ZEI9ID7ZY zEBeuu9VMmhsH}aBwcYC-VvmBY6Gd4{Vir3+3f|?-Zg0cog+>NhrrNu)K%8{B@7f+G zG;wtyo%=RBp_~p3ho7~hFSW>eSk1nrJZ<0v6LPeY7Wo25i2=#lsy|lyG~Umo-jP0P zL{vSY7%ryR8`}NDD@pfjh2pAS+?i#e_;4%_^(F87+5gBN{dzBkAFJ3yOLdA!vUTlm z%V8A7%e#yB(xu%q7#SHC3`L>KpmN9hRPXmq3bt$GK|LmSX8LoLLPK}Z#R{e>aSS(k zVTbwnZtd+cIU|+Jkj)i8WhK=dd+j%1R>8PQDbMVQjH=z0_cITxnapZfxGuSdh!w?R zQQf~wTxl0B(kMhkRU|}dnjRg8Ep$$?$}XQ-HOZZHG(22$8?N^QX(v~R3J`yoixm>s zRYVK;aY;ZCwMPoRvEMB;H&XT6!-$CWg3ytYU%_)ZPpPP_2xiUGkT?0S4DV0nM0z<1 zOO-1EDY&*;RaJ45dUG-J3+X%YNps5^y&qa}7|i8Au~iiTn^SG%;t} zHMKoqSbzn(9H}9ckz-7570CSUd1(aupXSy8F)KCMZt&W^`{-jUt>@sdsDnJI$1YR1 z2}Qeuj3#BS_?6c_Um5Y2tizO4Ps$%@?X*wj7w`m*>4p9V6W%nDnr7Ua@9Vx_e-EF_ z%#WV3&^Z{^<_n;dS+AMhXp8q(JpQwcPt%nIL$~w>*%%?+Jr}RXNX7cDSTbfQ&=grf z8qNYy%78w3L6;X~i+wIQRogWjUHAC)PBd$9(m(Ugo`FWLA?X1djp{eWCcUYI2RyCo zQ-tL_3hGOM>nO3xDzuRv$Vzub+@|M@MXe{A3zYr3GF z)aV$}sl*^0F7Gg0v?-K7UtuR9s!0K$_ki~r{x&7gUU6T5E*s;d7(UC=xK6J+Q$wgg zO5{!Fc&As`J%eCMgT6yB3}~?L-2#4SB4S)eF}MX%GFv?zdy%ly>-eyQk}TjvXQ!ca zMeHY#;(2W$NP-H46pH||xWl51wdTL@f=ha`-_&W=mKuq4GUz@As3xoPY?6}TT|CM6Vgq(|%7)#VYCOlzLq*9X- zg=L8go^iE4P6q?N#pj`g``J{9SYG;uL)2&Sw#2skMX1Fom-Wfot8n?E(%vq4#rUAz zU5C)FFi2mFcYz6e1Bm#c8f-1?7i>vi^JaD#Elam>7+cTJYV3gJR546LEV}&ut_dZ$bt7As1 zPcdg9wrduq<>vKLyB1Moee@|E;8QX$lK#1N>+2$=MR{wC*6imGQefz^Zr7!%>MtLj zTURk0%yI6BX>-;&36M1@ab1NVUTO}HW7~JXPx;pa zn*<7h>XE?3CH^(F4I^9t3>E-jk^oC)0JZl4|D^Y*0gg)&4g_+SD@QB=nfchni5&(& zQuaP)qAH(cxIg)rAs!E$XC7iM;qHTGLs4eQEkrSxJ`v)B>0= zG5z2^9&jIc96%Y)2uP)jeQnvVZh=ri6Me{XVAb;h5d|&_eYn5dBUfI&CfgD8k`{>$ zBY7j@ZxN`|V zboo99Jwf&2CCbHK_vy2xP^Od37gM}yy*o08UX-BmSd+|^4%kx z%4EZh{Env=ToJnRr4=l3@BJI>hLJr<^WX0ssgAFg&gK=}E(6Z}Po4KAY-}jW`rBRx z-(I|7TG!W%?ZU)}f=g;qkMKV#e(A&ws>9&R?MQdGMeu!lmV7LP>k_3~ozo*0SlCXF zwjd3gJjWLSa(p2!T-qSJ>_*rP*-1^5J}r}YZGL;;*3&vsB?Gbpdt^2ouu~REJlvV; zK8O-fciZs09r-}-B%_30pY5;x?9JV$la~4A_HAXP-gvMloiJS7of%!u;WeLcXZvX4 zZakJB_`8(!cBC}l5POBEZdRFbcN{Na^pk)hr?k${dbEu_jXK!`3UCcx^Yk1(Qa2ut zVvoC}!%p}0l0P`VdC<=l4kdfK0^C^nRClvMnveFS=F=zMC6^y7Ue+Hkcq^mmXn);# zx!=%V%fwkzb*{rpA?UAZPH?&#)>IftdS)C1KeR4aB5&3Nr{*SC#7i`7=H?S*(HO<1 z`snRyjo*@RuK6nq6CBD-P8ev|ZxjWdVcFbr&sweS_!VDGtl(AAJNYEULiUN=dOwN- z!NpAGadK{Ax#QUS(o)%U<5`8X0Db#UQQwjtg&p{v9)A5}d-+^0|C5>OXYZ}=9h^~m z;^F?=q|G(ADi`-2jm~gO_ez|nO`ksO0^CuHus$H%x z8~=4*9o*sb`SRb}cw6qcJUECVmT{}vtN0YBI-Lew;n*ch^K%hc$uRwo_>aK4Se!vA zKh&vI0eiv!!OsR?@gDxpW$}JP+~8|)n6atWS<;7W&*46GHEK+trv~M*!fT)C2g*jU z1n290q5Z~|fll6`A+R7eck_e)?kY(nL^9599KKRtezn2c`I_ICUB8sS;KPMeK^@_q zuFR$!1KjV8{iW8K{3JeoNN29Pl9xzF=#@hmEnjL}P=zzarhM}xHb2rfbKqKI`~ILw zIvMALI-|UCe(o2RHS47?WLDT>D?QkM2rOHn36}J>8Tdvv$55<5LdolZf-kzA3lnC5 zkm|)&C>I8;sc_HBZj=eZiMs@st#If}PJ4MJw_nS@SLO+zAZppK<^G|tWZPR9^GerR zG8~VpN+kIJ`{0^yUF-4VHj-ec{*I`D7G+|?4yF;E8b5!WKOiHCm!1Z$-hVu;A$@*t zr%Bxw#S;-~L%GQe#C6U$^sCJI?H55}%V7GEQW!tUi)8&A-0p3>O?6lN4cRBlMw3D+ zOF?N&#Db#U8}e7({CmqVAo3}Xp3s-RYh0RJ_MGT4V``_A_UL*%cgr2n39>^FIHAXc zN|e`o$&~)V)lA912e&o_pHYKg4bnzj#Va?2e#o(;jBRwSp0=#ScXjR5)jky>GyWXU zobM|eYv@7~NB4AP*u;Jvq4oMhN`hC^gy5ZxEs)mLlKNk8+b*ts@7Rm;O{M%{hdZf? z1EZ|@CbmW&HE%nhrW+S~(p}&vl+;bz;Sopmx=?DW@eM_Xduvt8O>X;mw;xiTbDOom zesJNAh=qjL`S?o3z?W)&uk|uh8=K#XRR^?$gewh8AJt)JzkMQqj-Qe7sXKbDvZrUH zV;~q8y^`Ipus?lg``FJO#Oiz?|*%gKhLM)n|KYS=||H}Hc8%9lK zv9xz5z(S9ktKLaaOZ=lnpmu*qo_Ui*Toye0==@m1e)0FvtV7W}Dgr=v7U`a>&6KQ-A^o-+K7$3Xh~HK&=H zJ6im{JIC&lS2P6afx~U-Yp4%O`XWtJa`lZDk33YCC4j*ktM}16^*ZzWnXMaLdCkhV zkdScaEs^xOhXJ<}$^DN!Ra`l_25P3K_}+Il@9+)l!1VX~J}>(b^~ZnIz<#tT@=dW1 z3ZHp?HlxUzBGgEPoKKzg1Re9dDj{zR>t1f0?4o-Tz_V}pi7-q!KP zp8w*+OqNga6EX-B>)2m1;ELvt7u?jPmV9le%yuC~A(&K!H(WasQ2Z!G#hXtl_y z7#8)XBqOQ^z<~nhqL#OQdc&alPxWNRqQ+qsuH?=$B6Wz+x>S8;^4Z+0#WkE z*E;k-2XvTB;J^m8XCd`Yt300LqbcfB>|5r1S4ki-MLaY36#`TQ+Oz`R00+{zrK_^< zo|ddKh8ZM%8Snz?Ivyr6hH0UuEk7gbZalWhL=CY;+ug6{ z?Q%7VsXiC}{yB=-xlv==dacm&O6m9jIvD@Eaf|Av-QAtNBPq=1yjoSS?A+_$nbRc< zUkm|!XoeJ|;NdN`dnIhzCb&yF+qZE36*CX73-%14&F&CVZk2HSr@*;3#kZ+Ef=U0-h+&Oo>pWBB~+L z6+!ZJ{8)q2zy}8_;VsYEgW%)3Qc#p;oA`#L>UJv711X~>Py)vk2hsTM_@mfB z#sxwipCqc}!G$U-)ql3?MiTI+(#SjCx=<7{)C4gL2&f}kEO@nJ5w=;LnR}UXE~h2v z-s@;m{u)Mn*`OTte1vBE?LBy}*6bX^gB2g8R88hS%;%5J4O^g+5#8ab8!;*D61v=7 zdB%86uytPKH=}_4PwtW&u$32yGEO&b3?1MQeOP|Bh3Lto?bWSR_nwmY)6q}li1Doq z@?)~^ut_4@H5OwX>nZiiA)^qCJX+mNEc?dew1TS=yw8nOz{*3y<~Ud9Mcu%`1OlyK z>|~Xrp68VEd``>1{oFBye_b?>_%fBx^XQ4POUdB6_1nago%$Q)Q5CtZ|3sH2Sw1c1 zxEUIzmW3udSq9VcVS`IH2$Q$}WEb_PUEp6edvu(NHsED^nZc)$ z)0VHU|G?>k#+Nt;4td^hJ?Yy*9s*D1Pn5r8pbkxlP9E{440NPlm0H| zUctN1_mJ`ze2W5`u+qyLBUl^6m5--}OhBbdDp13`eq zm-?`4sYh-Z+Q~I}hevK%o>ouoR$6TD)khgef|bn~n7~GPb$R7DD5OMI;oflrr%Ama6=37%*tuYd=^-KnpC!R zWTvMi00AZbFv4ar%*8Q2o-B>p4O7eUF&4NMQDaZeIQcPl7!2H{D2XZPE%$K_bV{5g zyC~(+lrVMswz{ftKZ>tG8M#IS+A(Z*!?GB@v@!@=NSvIKItc4iTk#_)cefXWEdTA* zHc1{M=P+tK(?d&2d$D(bIWZOd3qnKD5&3f)YlA$KY6P7`QJ4BCEx`h6G!UfKNglQ4?sz| zk5@R2v0p6?@Wqj8-nW8ZiKxwWPO8FjFD=oG_%T1=29TDx6flEZTun|V)$}Sh`Jv&s?tdlz$69nA<9o<&`~_q$8B$QGcBlJqKla2!b(oq4N)eG z@&-_Z?wJ!ezosghE~*nNO5a80eAyrT$pGBYRG=k7f&D5vF#cZsMxgyU+E&Klt`v1x zBf689>}$_Cmz4wla0m8x#YvX$RdUFoZ>|44F=0&jvD?LYqA$%Y=+3wU=3;TWMhB zO9&4FNT7P^sr||j!8)*`@pIa^xj>K_0zyYw{7i?|^KfCk^}M#XcEi0%wpsRM#pzjT zRE}yG?7kUrKoXn+Z*?XgdhQI-*}!oZip?!cr;Z5ajv6yTJ0;M^|B_0AffPl>=_7)| zRFLEsQ&mI5HKomOtfGL#2fys(7k`!)3!zPL9L||4e5&3GNI^Q1{E#C|EJLhqd8wu9 z2QM}6oiI|G`(hE~{DQjML}e!=%un7ksQyr@C*&gx2B$rc{r=7F#n{iqZUG*Ypzi&U z0p3eKYXJ5^31ysO$N{>$QVf1&sGE&7gerV%{>iUy>C_obFBOGfFx#EQ#EI5ePl_+x z5%;85SYo2s^7j|H>FB{qj*Z?WHD9I(JmmDl>KN6DO?#EY$Il_J{je?E-8bO!?Y&Ajy-`bz-ehS)FiE=HwZT7LZJvG&zgORCi(O<- zg@9m3TEP=i`|f!K ze3?|QqN8$hEW?OLD^|d0h11gsEB{)6C3(kvgg$@3INqJ&hpMWfew}2uHMI`R<|;fM_eI?fqXzr z_v$r35IE>X#5*4s$mktuxlg~nepi(wf@i*lEY|^~AT}GYLkSEXcu;*pEHa-(2FLNO zeenO~w`p1!h9Xy&aV|WdjIIA}DR zcnR^Kpq4J{CqXk^$*E&<$W-Yk;}XsN(%k+qxV-Qip{ts45D#ihefMhV_}XQ(-P?x8 z%G_j295K3|fWWMPN{8zKn;rVCX#7U5?PT#`jH)BAI-HL|!*KLM@O0Ewsw^nJ86Jj0I z!Mx@TSC?@vjgh2;lIJBV@>Vf+bsS$cuUgNT{!PRGPbI^KqGdaqo(Q#~*BG%&Dfxbr zOcq`G{0;Uy9FDte<3g_d<5>S))(v5n1i1(jH?B9KL4tI;g>a{x4oxAgf?wiexJKq; z5T;2%I8rd;_~kW6zHzIVx>E(Ue$74CkrtVeaMD=IOzU-AXr}+goeT5l?ZP;+OXoOI zlr4|)S_~sGB5V+c#@~(+hte&TfjAQNINxBGS15>!k!TLQ0C78FEntJDDbw1gmcD_b6C_O6 zq(Da)&+bg0-IbvwPy=%8TW5G6gbD-b0-GN7_8cYxGE4*{YVQsb%}QVf1!dc1>H#%E z2S-vs9_~?`Lzni1iyi7lm#gXK+CAU}$+ETK+hIL`DWy2zz>Mq&TlUOZ#xw(){-`~1 zAZ08@_RTFZ&ywe$QN@EJ0O(yZYqG?TodQ*rqYJFSS-k8G&?QYT^+J0Nb{Z7;Y03IA zV@uV~;4W8Yj45!xBiHBfrMJlMbY{*qw84Z6)YzL@vOKh_+!hkGHeDX8`)mjp6}=GTa2kL`(WN=M9Y-JhAg9R<3|YJ$_B9 ze1_sc#M1Vcq-V+ZLm-5bobPAB(0Z`;n-~bTZ<&s%ye1r8qh=E zbS8nZG5dL;-+tVh9@hr0;SiOR+IKNW2gU%b*&X~s-624IKQHA-byMhuVU{eVoyK<0FS$c~&C{7vyZCz34e zju_A+VBOOT9BGg37#=h(5Zgb>#NH^%YKk<>Bs=j^uy zZk4KC`PRVUPa%^|at6?#6I0|$t%Y}&%O@_vzUCQcP2@a8{)k9}Q^cTw6z+=XAa;To zaI@TnIIc;)vcQ0`0u)B}Qse`RRaFsJvBDhxtGM`ot@FS9K_XXp?YbUWI8+_YB6auS zPJKDv@GUa$*N8j{84jgh>}vAw{`P+oQ#8*mxP6RI>UP!WTYzfC7^=W;CCE_`7y$R{ zG@7LbFcv2PvneMQw}rfP`34}?_!T%zlO7=mKA^>XN9{>WisNlWs*K^whUB`J=6CW9 kfd9`Op#R(qI=@W3bZ{L0$VxQzheWwSZF>wJ3E5kv$Na%j{uv4lO=n9 zQbRh>2!^AyHW&c`>*doAB0_392?7Fll#Ij&HTU#`B@cJCsT;`2k-&E$lxG4j@YHI5 zeGH>1YNQDar)5%GgiPdaNG(~j)~a0n&|$5Z%B^E#&4q<%*dQu@5;ad!;6X-th4Re! zw=me(Z&4=sYl_+18l~mYvtu)-t&t3{Wxs&s@?9plha^Zvw)om4!WlV1MHV#+%`HD6 zT%2q%QH2S|%289cm@ltWpE3XMC$T(ogs~EJ=E)}TmNc(8W4_5iI3e>Txv7p1e#D#mJOEvgLc*0D#*#nxxLu*j`qdQW)HUMb95XE8g}*=HvMLa zW`5aXyl?Zh@>W)ith)7oAXH|wVuL?eFUaFmzOY#b31kNdNu0Z=GhrsIcej2xc2(W8 zl=1z9OFi|^Jf@R=8a=8M37Wv~ozy2p3@Oo88BWmUm@i4E5;W(==xnhsnA=H}+69-Y8oQa(o-gr=sZp>+Pqj_=B{l1W`t zMoi@k1rxVIVY_YFOth5<-2pz(WGedwKh%3_7{Y^TWYcJ!4<^TQ9URAWsPKc`mt+!C z5`HH-)4pi)=`wAEGVPj>`+ILi5x8?ljN9Eo!(E-9wKY?zW@SaS-S7A}kQ111H#YC} z5>CEaLjO>`=YGKpx=x@!@13R&%;(@08d_q+*Z#}U#TqH(SrCgm(8F&pO!ubC z5d^PcL4qeeFA;2)niJ*Nn!YZ#c}*57QPmr{eEHtL7 zE-TsPQIj);Al+Z(o5?Fs6k6G^Wgn?cN1I_|_wa^em_&7R-ZIPP?wQdKn>$?EUxeG} zEZBK*xIT@U*dkOo>v;6rt}3cjjlWuE?qQ~$es5shEq?L+*VG2y#mkZyPb?n zQBO5KLdhM>PS<}&Qt-!5OsJ$QZejE>FfzUp5Fqyl)a7{D>qvct*KVFG)BW$am(isO z;jfDAjQewADcf^g4hxaDBKl25867J@o4VP^RYP_if;06m!}?K-kVFd=GstEmx4A5a zl6SM*R(#x|?xHF9yAGFHNP!(Q!e%(_QPug#nD0F`2FpaKMHoRGL0cw@Q@!W+2U(zT zAk)y(7mlbm@9i3g%3d8VrM$C@?z?;O;)V5jz0FKyOA?v&T#gFU3o;$n@93>IEGr0# z*@RW(+uQYqKWub|ny&vwFfucnLL*717N#ro2mkg)@ZYR|pQ|!PHa9n?c<&sBibeFt zeP7q1O&tz~_#<*34)*ucFf&J7t@u&<|N6xl9uWbz*8TVqO)8qg3Zl)JPZkk^c6AjH zfgrK&OoK6tJA_(69A}2VgXE0bjv*${ttjDs)OM9D;wM~~>B+Rz;$}IJz*uWD6CF({ z!0CDZ+4Fq6b9?+J|6W;*wP1Z!Rjfgq=gIk`l2TM z8njg{9Go96>9Q9w;qaHn|u&3qbeh_1E;b&8A@9dPuQ(TtOB=W(sJD z1ZDl12}@6>Q1z;gjw0uE+Dx&$b>kVZyuD>2ypu2s2EFZ@^l4Wcj689gQ82jld+t7T zwaJ*DfXPFdQv!gDbWn?m>ON6v6gJPT_PX4B@W&Lm<;ilmtR}vC9kn&w zJ`4=VzIcJ;byJ|2F_0lPUW#4w`&#JnTqp)ajLQBqhk#1&yP`E_LW}nI6$?lOnNa#gzs!2scCHr zX*?fnx&hUzp0jm6m`_wyr(S#S`d->^YkgpP6ak zv#{`|14;sorX0+AIH#}d<3FgKoNWTrhvHIS&U(3JF*Q~&1`ELf&y)(r6@BtxO-^qR zm^CZbL#cfcjE-TQLn-e9p}}D*^4&<$m&e#w;_d+4mio=qMyf=`pj{#8 zc^%YI2)f4th;p_F9#v;gUnYrU+5Pj)7k>L?&OGVlZv=a0Rq7=CYDLM|_!Xcv z*#PDP4uC_V(qDy1XG&nhfp0sZW3{ldQl+}Gg3+&T>&ZhGEyLJoE5rJQ@wK;B+YKYa z@Z&TpRwN}d@#D&Ae`!eEe$5>RygDa{W9=vVzfPyLOU9Ewp5z?_3 zOhHYym?J!Zl-ftx!I2lrf`sGQX!G8Y8kSj0jo5mq!{>Sw@nR4lJ*vT$J z2GC(Z;Z>BCKZuDTH7q(|0SM2_lXf-Qv$&}DgGIXwn|Bcp`hzJy@GY8!_j&M&kelg{ z8#I{*nmhu0YZWI6LzCMedN3l*sUd#RIM8O#xB!Tjn>%r;RI|nx?zuUf)<>V^-6f7I zq_$-W=zj`7*dke{C_yFjMg{ZcZ508ZxWuW1nKxyN-*a2^dL(wlmq8N)3}XcfBIjjiJ*IISAdYnni$QT36DU$SE z2!Leo{^KNfJ*FcfB7_bL6I5~t$ZN!Ib4z%p)Tk;Iu#K0?a&rS^zQ3*Zys+MP6-eWC z!1*i!0B^xH#-VYM!cx1B0I;k)cJs$$;<)4WcDewi37XsyG(N3h0yx8!)xW|ix^H{x zicwKfpRi3nl?wwur>H3G@?gQVX3l*rinkU_b69*e+qsi zB*bQ9C;-m1X+2)9@q9wg0~Uaekj(dmwzBP#tj+7nVzNMCD#HDIcM7XzN%3g8Ek)R; zc?O=K!j#}Vu<}vX&@kU5OYx>OK_x2kU#hE2x%l&yIYG=5WgMdAkn#oyXhH%34h3I! z07HKyh=R{iRE{$dv+&svK6RbTp|-a6`o>1iBqF+@tSlz*F3vtgB}xiqGc0BvmYJL^ zoy4Y3yYw7b6+Mf~6Z~`I)Yrn+|1e?Re_FqMqSp2jM$zsEP5&jv)~&Rka!J_QS*B0I z!}edsx!5l}(9)(!n3adM($fwv9j4LJ>PoJA;ggc(x&jZu{uMu8G(-)Vn&-jYHjdmu zwM8pH2$SZ8d1kllf&52gTs5S}%d2W=GAZPFRlz-W9;Gbf6b(Lu^7EPaY;3?tIdMpEp7Zq4 z`_ExJJ1?*f)+sRN6MSVIUCFME4111zh+M%5C1_dVCT=v4h?mU68Y8@Zdu+j!ua^Ey zHFG6NrJpw-A3{nSPfxoRu;u^UC^M{#D2^C060{f45P%Kw3wV47-OkOCEr#~#+6&2h zJQV3$C;{QgeCr2pdMq1mn(d~G<{xvDP38$IC&hFt&05-xWT`GS^@Mg+6Es`H^=t(p zDqn53U2YNc)x4dkeLP_{dxm^cxgQ%t3;-&@B;0hBz7$+aO!lQ;E;vTQH?6)NZA ze?OTAYd(FF2_Afy?s#^Q`}%{|ArMg!tY(7qtVW+=r`Ug3$lDO&v$3ZrXu0Rugz6s( zl15@MB>u!!JYIH3czg!JEk*}gK|Dr09n9hLf%vF#J0f@Pl+@1^STLgBL z!W3D?{QLjH^7OnQlKZgB9XHIn;h>UykTbg}tQJ>AH&JD`k%UT!3@t>=B$EfP2tX;cTVr76%K$9s< z%uE9SShiCWzL|K{W*3e!=03 z`fxfjK~ug1V8QMt?&qKR{s@SvODv*raDI8z;@S_fFcX{k5?>-ps@ik+6yUv z^|?kxw)cE{+_C6M@uBWbHdUcMPq@ef%v{aL`D+QF#seYPpOo?A#ciI#TxqQqA~c+@ zUWXXJSGk`e)<*a-@^eZ|g;a=x@5}hYF-AYYZqrZCjQjBzn1k3mzA9#jopL>Ge(}5* zcijhNd_7mv{W;{v8{E5Jik&;>UeAk^{)v6&-o~*65UBQYAwZOln zp{=!%>nP+7OFaIQ^IRDAElZ!DVi#*H=pQsFgwhSNSJ~Z-TeClqM)Ekpslz2>vN<}M zV;WHltDVAnLaJO%7WV!xkMQIBGsmMg77Ffw>Pm0%yUpUEb~~efWPvB`4Nen56wn}E zcVEZ9UZXbdLU{mBF3G!_qUta3tR5_&ja6vNYNciafjFVT1IoUZUaaXn*L62F?1yeL zyoT}KdYZ}8lG}2VeI!THQ)MKJ`O66dnjMrY^@%)fMyH^*LdNXeOnHu>)Mp=4?`~4yoB&pU3JPiob4EBsaEb&8AqY4M9a|bTgEqOc@kpp6 zLzwbSB&yEV9}Lv6ErW+w9&s90qSt;6HXVeo9AD0RACfw~$Bo&J9)7#(5R)c4$%!Fa z5=Mt$1)+kO7)9?r%=nigdjr`577Q*WgO0j8gTIxri8%*GY(#_uS~v3GJ|GFul?B| zGNSJ}hGzQ6O)W@FfqC--(SrTwcL^%hT{DZ|gG~Ezv`1kUF@58Ac4YUrGZgJkbU}AY z>wPci9q!Ah#akLmJA_zo18W3XZmvWpIFk@WNK7#!L4=SNNRud7GZl>nYnIZ8JW=HY z0<@Pcv?SWW?{4~B@Sa0HrR$-zVl8k#Ypsf&7Uco|t1Ga# z`aVLZC7Eq&Bz)no&dOeTEqAgopI@YX1s(4hcALgj?}U%1`)w-R(=mVNZ@{Hvo*c!1 zbS3-Y7U@7q&ip}}^gwkjLFBw2~w4 z*h`pfK6t&x-C<#IC3oEy<&Wz~fK$rVu)_|% zSWd$8EONX?@SaRMJTm_NrPpd_iHt*--2w+7+rKb8TA4Di3}%DDbw>+)YX0=sF(RY_ zm@WB{i#4ag1l0m28h%A-^ruKlrqPxJ?90z#@h_A~$}*sxbLHVYOe%qh^+W@@M-oLc zp#J2QM$ln?E}7_qo)EyRtC@W@5sm3sL0fi5e&@}nn^teS9_6Qu zNCA%Vb^_pAHe%(X^b@BS3;v+v>ovzgR_~0d42-p`^&78B*lvxWq6zu;C;X;;U02Q> z_hS~qSB6oCG~9A`Y6n?`07i9x*)Svo#*0}w(OhnIW0<2%L%&>9ep~-tRIXzS9ey zvaJU>`U&6DB|)FnVs1@^@Yhcy=Z0-ILMYURwtT9T@xY6RFszXtczk2r&< zA%L~Qn9em?7)AMs~jL4eS6Skf#y%=SfG6fLKygict$ zApYFn+Z56=C{sQyi&%uvU3qJIR~j_#Hq9$}rA}nea%3X_BWUfktC%<7axn|-!6xzs zV~l+EnLHX1iD@$umR<*Q{@)KI#A%huQ#2B&5w)AWSwsm2}0w@r21dc zt-l3Bz8KRyd+vJ*+v^+s+ANMA!v~7SGlhnZlRa^>@D%E|vu`q$X<}iS=9JP7v*M zcpZD`3_CJ5E{?H&YADGM_`{c5VP6sS zG_zkj2Qf|SsjmVk|IAq7FINn`AF_eL8>nm%>eV3{f=<1RI2a5fB&yrM1m zjM%T=ZRvLW@Tc4M(i`Suo=6I31{K>sOsDR`L`Y#NyApLX-^C}Zw}H^UUMc}PPrg8 zhod_V=gzG!ab;eTYsB4upG1CdKP&M;J*og0&XCH2SZP3jA>DJ6W|Nst6XC|I(a5_i zLmE(AF0bWx7r4s@e%F7bfvCZHKQ}LWiKJT>i?`};F24NjR(C@BT+li8$~en3l7K`n zgdL!DQ?rttl3moeiO2Wn{P+7!jLwYraotx6ITxn5;$o2@42E*d;tU@R`SaBTIiMU{ zX_Whok02_oYo^2$*0`H9E6WUdvV{YWr>B2N5fO2JO?;#m<#8&bLhOS>pGBe7LgOiXox3gm5eUDk zyH-w{jnD=CVz!v7^;Dd65z@p)L}M?s5EpGD8F9}&5>$LKFGb0y zqZ`U$41Vs90`FS|A{si}gSjl<2tM##wjALc`sdAmsFlnLn@L1g2Dul@Le7A!n~)Zk zE0+uERYQPV6sSgGUD7BDv>DtSodeJU!Ufp@O2NB1BH@fX!Wj4!|HzFox5XIE`bJWY z%?zHPab#j}h;;uZm^h9Fh6w`!{01Lp0z;c&t3<>_a(Qx-$Pyp?-$9Q%uy5{9+PI(V zazrTo7@~c}k8j|Of78oO=G$2-tjvMLHII5Lw8ID!rs=-$l7q#9)PHlF`r~9$G0?vB z$3VzOg=JOZYYaNKKUN zzM6^lpA$GjY>!ruMF?kZ0nKm}_t_h+}>%5taYEjf#LJ#8G-LR!@VQR@=Jz6f&VSgNzU&ZlG|fvNlt zP{BKZ?x%n@KN`LI#Ep2S1q<@mEmjje`y+Q{4vA+!tJ1uf#Uesi^Mm(pBC&aSf~Z?L zLfsxb=y8X#J`8vu$GPmmRzt}hj6#dZ8rRtfTn;->Exz^tgfPLVU%BJt{|zlPxJ}gE zWiLM4$Td-iec6(NUu@pLzTa7kvb}Xh#$~_b71tQgJ!Pq;J@X9`DS zYr>KRGPmA1-iOU?l-tCSYfMuN?y{9ArDQ!q$7(0Wb@>2raQG}+2y52|%dE8b$*r_T zP$GBSWh4e#P4t{0zxIyfrOo{eix&V+X4LY6onUeYkV$LNjL!O^#=wqY!$BK4Hx847 zEf)B0mKgzM@Lpx8)#Ps0dTCDEtk&U))Hvjg8vvH!wV$SK??~KbnJHT9>5{gH?KkLk zHQpUX>^LuG;1(j`UWT+RpC8G{LEGwxbq@_Y{)8jB8fz$8nm8Gjc#jei!NUH)o$8I?oZ5h$fG zY<$f7C^<2?`s4d*uAsEjb(XmOXLcUhvuV*G-%@~`s=fZ}NG6i&k}dY6Q|9hURH`#s|7Vai9VWwNkC2SS&pZYaQf*pK=RNtQ&muN*9e-R!j?vyzo@ z-g$qRw5ZHOm$&{r-!Ccn{Gg$X4xDub%Qmw0B(8Vo;yc20Z9XYSPz;Rv#9Z;JpZ?ui zV3=)-Y2ks#q$RbBtBk*DT&Hbp#?x_#C^jMljo+8&;sA?|ndqn=cIQMuaZTp`LW;Ad z-Jx4fPI&8cT-06rV?vhi1=}~NaHDI-{_u)GsTt$$uukJrM;kIOjzO)|VE*5A8m*q$JK{ z6a@EOPo`Ruww~$&tI~M+zVXp_d+DwRJ%o81`Z-@l@F^@15nOK|JrCG2sAMPpOEw#1 z(6)!i&b~ZCzjry|ql9Ik50|EPnb1`PVKBL(73b;ac$865o9_~}gmr2i$MKTcQ zCUM4ACwx~@FzPaM>pdES`rntgnC@92!&5awSWWs24dtx9F5nB%E`Oh`R@=uM;t5!JF6GW142wQ7tV>FZ2 z-}))Gf-kg>`^Z*hP_2_m8d^byJGudzWKqV@Rud{spP`X%pz+R zJUVpA%Hh7-I~`X)d%R}6to38?uGBYgUg)5!{RyJn;So> z%s_i_Rfg(60+n(s`LW)3{bVJSl4dz!ws>(Dv`&RKURd}k^t8HqBJYA`^zv!o$}II{Prwd+1abPFu@vATR$7nrdk}41S~ca87&KeufVg_DS!3=iAwjS|d&7 zjqiO)>Mew;c#Gb;QnHy^6i}A9J2;@2DX}9f;cK&fBUOyHwJ4b?0w;gGA|%4+*YVGU(RHDINg~NvmnAPd~`dA0msiG)b|H<&1Q@S1;`1_FMKha;{i087Lv-|C_%K|C zQob*$^Fcvi7RaG<4`Ms|dw@Yjp#df0euTgIKeoC7vV_kVKzA zaYt_)BmrwJ5)E-fN$^jSukhnAASnKd>W18jn4v?q3|dHh)=S3dJ+fKX-*bzXT3Gd5 z#Ex!0ec`7rXEndP@-u@+BR^lR4xg!4`Z=@9jvtCVB0~Leji={EnU_EHn7(Yi9VGNS zvB3e!mu;o6i#_|j8@>#RFp@Ag^nR-DzzZg3Iz**u@BlU)B6XpZ0v|G!$)2uuPv%Up zva`5W6^d_wN`idFR2!YZaKv`k%!N#}XcJk=uCCWx#|9_TL{xkknF72{Hiw$ewXaYa zC4MjL-{$xIKO79NOU z{tKv6M9}2B-Sjz zRfy;x8tu+2pnd%j62;p!mJ^#y((~O3V^ea<2=$K<1w9%%X|PTa8Yly6@wqoWZK?i( zFSqxf^vv4!ze$2b#k_hgrSnrpoH3bLJ&60+@0AY*nac}C>6)*1;pi%?XK!c>4yhK$ z)|fVGM3Fw$*pKB_sw2&nk@F3%j=e8yZDD3rB|H~GLHRxr!tEjHeZNE{Y#@;M^+3e2 zca#0*YFcE>{EBVw#OJQ-e5YT@5A$Fl^fa-Ekqmw9nmk#kt)Y{z@{?jRkvTQRdolIi z3Q;6zC*kzW?Rcn-?^2)r#yqI+QB*$${GPIGt~x8G;XW!R0xj;B9acB5U+>Ujx;$StXnJ#eR^qc5SXp+(KyoCLpx{u} zso%@yhxAcKc;C>;yDF+--jTw;nCh#l&Z?0W+YobW)F{@Sqtl^5i0xgfn_H{cf{(|v z_=phkVal@^J+ODIa-z|F=iS2V^B=O`%tj|xa4D3I>^M{;=;f(bpBvys|7x=0YAoGW zH{@`_B~dRXbTE%Jj4t61f6+P}LM=}X=_2$PT8JH;=hofn*s*0>YELL$Fnn8A_WV0d z{~K}RYcrcykvNpCS>hZIu%M~ekxzX`}}jJVD)e^(OtQ+vnbChRpMDK5Upo z9{d4qel^g(zXvzxg0VeA=|6q_$~4aillbtNfYMsPs8u51*I-a(lhqD*e7n}F=C*)#`ZF#m+qa#s9kakRhc(57T z&bDxu`>e3YC?a#qEJ=Cl9vw(RUD|qAX!P#jzi-G{vb4_ll+{v;BV3BMRazceWqn+R zFaC&9GvCHJh@^B)P=B90T`M-TA1Yj_a7k~lVWq^;enE^T4+Yx51ixUI1QpsP32E==xs=M3VzAdFz_Ovt5BE&RM7WwIs9)xPe?}-YdS>Ll8ATliw%SLrB62kYn%l-7TJ;@W`Bh@mXmm39@^z zAXTM8P+jX7BC^}^1Fy_k=HRN&=PQ_?py_GTRW3r^9rL;0ASm_=mM$A?;#+m5jvW2^ zm4t|YV$Shg&GmjP^XG!9^pZA|@up=bvm>LxybpmvVKn=?tUV9+ zebPD@T?d>~C?H180>^;9XP4dP`im=xeaO^!4>v8bd5rhg;zGjl#-b&#NFRm%P>Owg z_^xKVs5D>kLUFk!bn@zmX3p5>f=x*tb}X|)mrl!dFp0LH@aS)VtR!Fk^?mj&`2Hh> zzuGpyLp#^(nin;Y+t68X8HqyiT^&Zu9qh594`h3rbU!yYBuArqS-G7XU5Llz6)Mjb zbEn6Dj8GJ?n%E$z-MZfGa8XYrs0bYR*|&hdZ%u5d!kYOxTk}Ki6z70jCKB!+3N1Bl zW`9zt;fSyHeH@HJb65IocdnU3)zqq>yRUC=?)Zzd{S|vdcX|x=tX_Wq?T{ujo}$2T zkgQlro>}mlAKuTmL&qu6S~<7li&%asDd&|+crWhKS(-yE`qLIC9&Hl0xTrwr3Ob;R z=XR8+c@f?@I3Owd=F_uKEEWDL-tT{Xc9Pp-RX&MDz(cLR)NZ1GFxEjsEl&m4>XX&u zguKm&9)DPe=QS`22={%~2N%seY{Hhol}q323&qc(}K(nN{@_7_a5=z}1ko z7`h)#VW?mH3Ryvnv~d&r{g-8TLn72~JlG0y#o6Pv@@1t*f$K>>%ZlYnk!lceys?k! zN=Z2oHw`falDhGcB}MVo2<91u6XYe0X$z#Qi20?3^*&gCVdODY|4E0)$}gWH6t|?H z`+cUCHCR7Yd72aZHw%5fI^*yHRlY(Cvh&cal)%CVZ$-~CpNu+Io%D4*M_@GSF!RiMbC$og{z{4(Ry zKzpNTvjB3WARNG@SMlIN+*1A!O}5{cPDiL2t9AIQ4k_U+opNmE?}yq=0$8>>N?CVfd7>E zRr(H#?9W#*K40W^E^4Y2^358V>VBqwrXzb!p&*rX>hQTZ~(mo|mB#0L)Fv$MS=HLzzXCx&k%>R^PFP`ep^^T6GlCDXTKV>J>Dosr~?Q2l2 z3{$R0eCH-ZiHlhpq3z;hY}+uZYMqFq;MCeHttND22g5U=jiS=|Rh9;eTQIKcCpUjv z>5-+G0dhlC!et~=AMhSQ1&$Jv1hSian1oINb=4F}&C-;=Hsl$^)X^c9U26PxP`~v7 zs>dNVWtjk?L1X4d3lN!0RqZg}V zP&jzVH!!{>yrEN0jg`NW+abPHg8`kOV0r{ z3@>M5wzx;}4Tf4Qtl6&b8#cNt#cUD5ULmpojw!s1fE>rP!G6J(Z{L;4&Xs~k04FU9 zvxTy7w%FOO^44IVLq*$6|&!dVGEI!{agzR_IZYP}Y?s~A2GB-7f+l&2YykfL7Sl+QE&3R%B)Vhx6#iJa=|#a z@yOe}3oAVTKsTx|NMR~QH^JEw5N+RK*;J1jbGILGNc{HY1cVQ9;Tth?kjKL08C{;X zNSX_}W$=X*Bhs)}5bHSR(8aLxwvx=x2CT+Il?9#rz=z9AqyuX* zp~?f6`lt}%Uf`}#S@r{~V-U!qtq{Ie@KpBD#jp=#;N9s)93fUN%iv`Zk@>GLs< z9PdNa&0@}%EWdx7qaJJOlOF-9h!T!Wik~Vp2c-TTjHJFJahx&jXrq05(kNlSE zr!|S}fUJlZ@6RpTpmQgWV&J^gPUX%#ae%YdWQda2WdF!%06@f<+d&c{3(^ln{y2L# zl9js(2oufZg;$z%gFxh#F!FW&o+~d6FYwgg8(gGYguT{9KmKEk>AyK!2!?l_`=9rk zG}+7SA{o(3h^Ae6pNf>ca5%IizRqeMH38*pWTqt=n+73uUFxmJ8*S+BDA(^kk&Qru zli9Z(uwWjh@~?n8e8c1g5~lX_7R1zMdtPMU=cadJ10*%i__X@iUi;5Png3ISdgrwN z8S5~UX!;OlM{O$q<`YWobGbb+pt{?Qmul~&kij3Xhl_*L#V3M){+-?LT-jc z(-q5obs9GBV4_+%&POl-7C2p#A^;aoy+vEAyYl^aV0A^%ZHI%4KWK3+w(lH1ee zEktk>h!C1gT{X>cNLFwe9M8TbOyy1vA*msd?o3rH1syZ8xxcXMgzILPhB@QM(UoBp zfNc)m5tHmS`%oUIe9`v^`*H!e?{w6AO6ICQxsf{Aj4g*qMowrgC)Ak->I`k(bW8-j zrP9#2+IWQDX|=gZE8$!vAX65`ctqL|CXfkqcd z;@WlgH+GLL%9yu1-0Zwa_HaqdcwKp_2(Quj-b~m5nly@j8ZcGi2Y$V@>6qzP?ju64 zSYqz&eT2%)T8DUz(6uAMpp{4C;w8bFp z=>A5e@~KKejUv1rD8800rg_R9{^zDZ@2jVq0ss4+(mz)SdUGDpoy~{IwSLe(0JpUe NWF!?OD#VNe{||h^LX7|b literal 0 HcmV?d00001 diff --git a/doc/LectureNotes/_build/html/_images/chapter1_3_0.png b/doc/LectureNotes/_build/html/_images/chapter1_3_0.png new file mode 100644 index 0000000000000000000000000000000000000000..29f3f18180eeb9c764b68bcfcf41c2485d6d31e1 GIT binary patch literal 10016 zcmdUVcTiJ7)bC9~6QnokEr1FF8hTecNVNchfT%%w?=7G-k&+-{s45`6NTj!*R1Liu zL5LEH(rds_Ui@C0Z|1!>Z|42==FZODyLj)qd=&X_(+d8q{G*#zA3h6e;c@OuoHJ_zvf@DA{BbrueK?C0<5?FEyeJkr8X z0s?&eRb^#8{|AALx1Wn_e^z5Ig~>S|LtB3UU||23KtOix1ptspG19$$J0y4gN9b$& zJ9$c5n>85kbZzi6=ns${_E~zm*&_?v+v$2emRPxzSh;754|r%EV6h0p0FK-GbK99* z7V)|kzCO+z$Z1AaXn-NzIo(8CIx3?ZhW+}lHa5w9&5!Q{mY2J?lJZ+fpeRxJ+$U=4)T~r+?plKeE@$@0B=xLt~r7qbme*~<=}PTieGVaG(HYx0mxYrPBbTA8U>Ht#kls)O2p;5^Ud=tYb`KV=;6gTrk(FR}R8AG>eML;+(}J6$jmZ5$CZS06UP&StJAxQbovu4>=I5z$qOeCU;r` zV(qgDAtV4%&hQieh5{yn2%rPP$z%dBKB_F)ud$suz%f78-zRkU0dVFF7h-hqV@G@d zm}q=N?N5V(6%pgH3!BUYZorFij7$w^DPUN;K6Ek?I8ZtOKg8)lZJI3PPyz7((EmJn zo}rl?J9j-S6|$uQ6b$1}HG$}u@S090LjEj5D(aT7p$f_yatHw6LqS9!T(J*#AiG^x zlM#_aWCM~x$JsVW~CE_A>l zGm6is1560+cgS=+PpE)8LYw)p*1%2N1ij&q=#j+04`?UojSym1_<&=e0J%nAeaLVW zi}s|d<&i1S;yaF}N3cS@MVetxlFtMYu3ZNhAR>-v2X%~->Te}04or4^7zS;bpgPZw z?Os4I0VOoU>tMM#98G$JM#v`M#EswrXi*UfMD`!d&hYlwfcxftuz^gV1L-dV)IKMY z41>T;%vi%F0+dlZMw^A>*)o!!F<38J7@=m|Moh__Z{$7<_Zgb^M$TsMlFFUVe$)d^ zs!yiAUEs)n_*}iwD*EeDq@CH{d%&^F(6%BV*CZA28qXaIN1JvXwEwIQf!)pWU7Imt zife6|PxTOX1^GBN?w*L&#VyDDIB;CCX+CoY`M?`bA^~i_@R=5b`_MrN<7m?`{`Wf} zcNvXmgQ<05ed8f35YpV^${XqCF`dg-2RyrKnhPM?ufZ`yk!B7oGXZ=<<2*yQ#IG2 z=YzUFSO=&bu%OUVhtof^-LCa=eD`ehZ|Z}{ak(cb(A<#4Zof|0H zPYPcn`wsY1>IL!HI{GmC?i@TB$MXnYg2tz_k=NZRA}`o}`Q~&mh@z7Vh|b`|d>#)zu#*xZ5D3e`cS62Z zd^%Y^m+d}4^KL`C(eO04`c>q_W$3E6ri{e70<=_pTM(!(3 z+iHEs9VlVOpCZ~ISgxk`qOb92G)WL^=hIbOVb`MebPp$w5xG8&@1*H_No66NKIkJ5 zX`X|>+XU2Jz)l8!Kbt4ezgF-ODCefFPgRGi0rqr+Gpb?7ctQPNfDL$h78M%Nu{Taq zUiu7l&=9Wsi$!xm`dV2YOQ?|Z$(;H=0EPYeOM(fY)0)W5; z)X>bF3fY$gjEb~Ftv?4kM6=3-RMGe_&IY%gt0p)=rjG;p5hk%&W-;NLjlgRN;ftt0 ztqg4~C)N)z#|Gxl+x=`#F1e}n9$8EEncIarlG%y^7gShzR9{aXmyzHkov!*LxY*#@ z34DswB1>Ip z(C+fEqq(21LQTqLmKUN&+30mz!c`zEBckowuAQA7esq*ItVLHB^$E9RJ#)ruG|nXU?-aqNDO%oFpogCe=WJSKD;3BJ$3N`ru!)&rN%WGR%n4J^czfkpkV&b0k7v8Q+ zz~RMK$iK-Nf@9DrG(NnD)gADGlv{Ld6}dHe5{X)IvjDQf*)K3+>M50_tjDBP&=cB z)C*Td*B&J$9zN`S;la}Xtzqy)3v4OR*e-;X#CeD#Xh9MEIJlmVJu9+z>x)$8N$1cF zW-4fnq)Il?1{ep9UlCOv2wCeEJ>YO>N-IU%@6ze&{!;Xpkn?dmJ)>a?I7`zmia`)J zjiRK#Jba!QQd7nF#GCfsJpe|+v}>SJorPO0btrQ+wzGRg9Ubh(%l3cxm&l?--_dKX z9=w{DQB+rD?@6SXDu9vD#L$0*^i-loEW7U>kA35y(?tm{>3#k3PV>AsWt#=Pj!S;E ze{nqFgBzzdoGiz2zU_^+@PBZa65D?eGHBSF7q9WLp`os@P;BE-qG7wdLUXouFkIW( zdhfT9Ff7Nsu0DIZTwhr9r~5O_7aSCcLv9_-B_RDMow8}pljW?;V99z=-m7yVdd9a> zv~|Z!{P}=VN(0y8ee|iB0+`iEa(-D+*T&BBU+uv1K3cT8>L0;x;_hsGqPrA@6Ut0; z*fX)I;hpw>E_U$r#D2;2YA?)`I-vt=6YDuC!p=IcYj0;s5&dmTFVDwIH>WN6$1JQE zg*aKaZK-sO?ufCR%XxJ*{hqGe%iEMil5^#Ol->Dn6&agfFC?CcUNA1#7G{$u$hw-| z%hF|GMNJ_)YI_kQA5RyLiWj8ARd%8WVVGr{M-n>n(%Qr4A)l8YpD*%#roMgUId2ye zAS4L^Na<$=cghQ`3l;v-!2#Ts!{19Wl<@BwOb35K^^kRN;Q5qH(hQ{9j z+(GL4JHNrhXfL<`eCmVOkCuxlaZ!y+3bZ$N#n?X2Qito(!uaS34=4u2-hM5#s;X*j zWhGJizNnQpNreTmI{(BwxRZP4{daR|oQLMXRJjA+V_^V*k%NI5!#RgF70OZyUQnAd zMXXjA_nb&KcrJdCyCtqS!AAG)f`Ytg={YQp2A~+_vD!2#QH9I@wPW$;nz2YA$_1Vb zs6(a=26Adz11 zY!^uyq*1gH12qRIUO)|Q?LA@c^RJAL(F5Jyf{u&^y$fXZtVshWI1ND-IF&_w1+{s= z3m9P!tU!?iye7O)8~Ut2BexGVhZpx)kI!MOi3c6eULRDeZXf-rgDLm*2DU@mdZYDH~gB~0+UanF)f$qCw_FJOLy7%9Z z%c};dJE~2d8U!UW6=2~*@ zu~GVaS=++K4<%QgEbr{O*?d_MOMLp5uBa6HQ6b@*ne^{FCPFdb7c(jEnqMe1=M6P#idW+ zt_h$;%C!C5d0o|AE*Tfb>!ReK^;WLx%g@G3MUeN3eDMkL`52C3iWW|#-!7OTD4e(U zyU3FG>?(AUR4gZdzA-9->2Oh$=MMyX0qBHmff3?B_1bTR3`z6%hDGo}QM-GFFp~dN z<@bZ>u^TNc>1R(|%QUL?tv3D53T-j{F_7K)g&4@g0TV!=vM!gY9i!YdylsdY&c?F< zyPyaZk)g+-aLCy;RQVZraaTQv?7&U*Mpx309mjuoFP8b&v$=g64u)fCI z02zS#eW#fK*s2^RHE?xG4=s)Sdy6kZ6$&L7aM9sB1b~^5HrzxSIXKZgmw7`PFyr;x5Sja?tb{xfqG$$ zP1nErP(O(A?F-MD*|Q$U#SZV*U#*y%?#X=Qz%6s}&OPzYW!Bq; z)w^`YDZ7!2nLE8<;qrHFiVRZzhOT#rUM8qIjE@aCtVMoZnc;pde=cO6T$lEgIcSm7 zEyp1J_;IO8%C4uEOz!rNgedVxm9<6P6l=n8d)}b#r5mr2v+&9q)hlUGe$;mjRW(wC zS&ai$>xA70{}_uj0H6GgL%k%@##+TiHaoudgTyB?UyECM!GmYWN5f8JZ)QYw?SFU2 z-#GhS5|hg0w0gJE^P607oL1oFt!{Q~OCYktdw`j$n4av{DVGu!C;oeKYAL#F{ys{k zslfxZG(0SO=omiexjIh!TAB6mco}1=z8OWD&k1q1^jyC2I!!}0*?nER+<^8rf1Zu& zhH}Q9>%?`Q*5Hg`E%UiZrO=;o#AoLo-MV$a&!qWUo$Rrbs@7kUFWTtMEY4H<7NdM{ zqjN+|mPP(gLg8Ik4R4`K!UiR*J7I&oJ3kW!R;$I#m1FqTmB|7EsxQ@lLw9=l2Acf` z>fOam?p}$(o>}UosDj`9YdBvN2{rldIn&_6KKy6W%WiY78ln`;F)WhM)U1EC^zlUc zS&VSi*Q59QZIOkDT*3SIv17$Yai_b(E(W_V1acf#%(JP^Utn=Z`G>Zj4;*BxYjNyB z1+l-&W_@Dot_(QOt`t)a2{cumNEPJ|{q?@6+PYkhX1_ zT4(Al!c*)WYfvBm!X1`U+j>vq2bvQGSYmE)^YYKtqm5SYsBAx%010%>ra^s!m-yMC z!Op?V0BjhVlY?2Ftl(C0WeeTzA(A@HJGQ`hJWud*sd(LBu8fq?7>;--h0WKH!!Y;0 zk2GM%`821(p zM^=qkXWv%1|G>Ry*P&Ih^Zpsy1!0;q+aRJd3$41@dOA^t1<`ir%%wj`I#&=V6o|6@ z$&n~=p0sVnw4R%_gKzJlAQn-%-f6(4(6``MYVXWXK^ z%J|t)>FS?C-(<}5bsaVDLK5p?c6IHM&s<&}ajcjaNkV2h7@OtaJlM!H9e#b4$m5dU z_(NawRJr-FNr3Av3){TGNdGT;wr^QMbN_SD$$#6XtNb%+z1l`izwoNclhFd|-%RIP zHdK|+?t|Gr<6{W7b5wuC=wVO-jrMH=7tmUw`Gm&&EABkK>fu$s3Bz>zrSz$ zGn`N}AoA3H93aK+KZx$4-f(fR;Omo}yCvM8*{!(Ww0zZYA#AluBPV34CU}OjrKUdN zw0=}`V33wdTDokdpuo6j{0sHj$YTTbccvLO)17hY*fDH?(sg0bHR{t*gI0G)+IVv+loY|0M*I_+hHv(dNx+C2Xk5D3W4e9|>y> z0wr@TI)lpJNs)>u@RUBeHuGCqACLm-iN*QVaNG%I-d>b3O`xa*akES8&4z2{Ey3htbpmB=qP!MGa;ggrUctvbj4 z%llVS*t2_O3!Lt3>#kHzVlV2X>f&vr5KMI~LwRmuhOe0?>7dshC_KLH7dT6q$34B% z!#?7qb%xL(pyH%`5coMEqa;s%&X3_}rH~lUxNr2ZT`yAM{-2OZ#qGn|;~?`JrL5f! zE8JlhSq!0{PRF)96L)n0|tU3^AWJw-DDk(u1<>%mmQ_XLgmU~M}_4JU8L(V+Q$nE zoVLOB?|pHvKh@CocSH`9-5U5R8|-+H<>-|wAvm8&?ucW3S$<_i>`?O}OHtD1;1!-s z2hToVhs=nw;7y5w@Kg_PN!ca@zg9;Gi=5besO0fp$sGlS=CH_)a>g6Qno@zk0)+t| zSwua+l||;Ijuqv7+bmv)l6s>&Tww#+{!>9Tk2ZD)zY;p@OMG)3|B;bTKR@DHGih$I zpm4&x*m`>ZI3kn%8NEQwx79H#F~CO}QOH5}=M8JN-L=a{J3WyOb`9yBXBtQKHOPX$ z94co^@Y;E?Vno%mgoyuTPINvU2U1-rc=3)bU)SBwWtWP6&kfKu;Ao zLWdFb&$JUr+i44;%qTj)jqJ7aiDBRzd;k0QYgd{7Ge*j`cb2Le@V!-$GtML$X=phv zQoVKV`%3j9eXdWJ3Ji>c0)Uk#US8GiG~)WR)u$Zrf1ZdB)x^qEkl} z`p?MeuH|rrp3|#5aH4vype8@o;^zFgE>-Mof*X|{wBDk0I@I%rF~tgp&UjKxGBNGX z?;kn2>gcvTtDO#Vf3eBH(2&ni32)ENn-QC%rl-+`lNJ%#^DLs&vGT4TWgj$O)~FdW z^BZ8>nF}9CmJyXSgNE;aiQ4W>71!G3*~yuQp6yBc%`{vbdKXr#|LSmp{{pGUu)ZML zaY^Y}=__rPfp`9JMJ}v8YT;{W2UlK(M%%XjDcECJ9nyBvFw-#YfSoAAQ-6DJ78s0o zgX(dyNnNhi8~J7ba*$*<(A(y8|4L)|c@2s~1OZZM<@m9DfR|{fdD%2YqGk`)v6$f{ z?qjs9O^t(B@!)Kx60^nxXHiO}jY!>Qhm2OlrQ_$+(y%ubQnL@q9oL>Thq8|ej8Fgt zDbIN4uc6+!g;h?-O%nCPiedP&Ry?H==Hnn%#1DsSZYdnB#E zOO5IG5?WrH3b_e3{Wgu_)=3R|ueZ*p5`Q17z1E7Vhgf4r4u@@>Gx z7gsQZ94#IPoh;}32Y&}6Mhkam_9e|}Xc#G`-m1Mx)l0|CG3{pJ3BN-ym$le-43=b8 z<5%fe;J7lJL+dfS(!eNq{w5lwg<8f!9-3B9*s^JU_NBXum|~*~n+|wXX7+(xHk8Zb zm~D*z0I{hR?j$&N%zC=JfqT4|L*`vtL$>s+eZzdcPlJYgk?JJ z`WpRbATFf<9Sw7;!T=j&lCCXhIz9irh$F+W{=rgCwXI}hlFMhBM$Zx9H}wjg1R;gX zb9>K))TO8%{YSUPdoFK`%jk4?k$d0pl|620AZR0C#vm2fIp*z29e+MaX;3t! z8*}ZFfI=*l-n>QI{?tDxoNo1m(6%%jOI7!C9Tyt3-a|~JWy_2gkJ3cG98r#W)7)#i%LpIvjLsmqbR;P~F69is4t_b;3 zjg**24!zAUZ*DQa*MK3)AC$>=|Wxf*(U~z{Z)g8 zMgu?9K+8eVFeAH=nvvyq=0d^2mkX@fGrJL7dC3;d)=@c%i%Q$LoaK!=>L!W&!c;%& zBhq|iyQ(V7)Ye>7bqskEa^eG4KHgE^Sbfib;WTHD##@$YVzkC2Cg>!{Dhit)=7 z;EiG*es+#xU9Hy1f*nfKcV?tfL`@+DPNQxPx_1*}R$&{Twtj}E{^C)rJlqW@(NKJ) z}4(AY`c@%IiP zqW~vUpIZ^^O&BwY={)7MY2|vY#@)27PbkVTrUkw6VY#?}7FOnx-%#ItqqDe&GvLaV zdfFJ}`Q5OKik-!RyAIc4f+sI?dtCG9;=PjAEi@S<^X#a9Rl544f5WSpllAXiUPAmmgv7sPaiYp(ya9$$jBb%DD(p z$MJWT_rBZC+k}4iYLPSz#j|?ynohnSedz8WQ7kRpy92+#XPK!?uQAWSHa~7%UAu98 zjchtA&P7rDGcw&}SNAKE*Be8wj*KF;EZqlr8{54doqly6H=G2wOSL53vuq^&)r_Af z?e&-nj9O{eapzS$N_p#?*Hm#A6uy~vMR#srhH$umi9791((a7fH}+lp{{7wh%5LNj ziTQEvMye}85>Z>7n2@ELB$46C6P->tURpCc;}7ObF31h(D@nb@oRa{=|Q? zJ*Mo|u3CV>i$tZj`&`T2(t#tr26>%Uv;#S|9=)lvi?2$|sONSCt!>0@bE#wL3lBN? z^4<$ea~jT-zBS6Q_Z0U0)h1A)|E~MtfrmALyG3~X;-#d-f~;>wojbJjvTs?I>Zmlf zL!rinvPza;Uhwqf5ILTSi$*Cj9<{C5YjCVGD3k079psN8cGX3)Ppm7wX`bXxy7Vv% zI0ncVTiQnsy+%6Lf;KiCPsipfcA7Z8R`9@n8JV|pR;<^5)XiZsQkbOk(+k`$V^v)= z33Cu{?-fH@ro3@@*YfuTM)u7pRHG(hBL_KQA>7Y7S(^7z=BcTaU^1HMO)+Idknmv`eW3SBfB^}1lnNqfGbSe9y-qc%Ya%GQWwvx%s zbPw7QCICJ)YPl#^T67h|bXoQ#v&u_u+b_w|u!=BTB!{U^4d_C~E@_1jO*znwxd`rp z0ZdmhPjQv92v=Z9ml7pmy`-~Zssjzq!9NBBW2?-BE+;DX)P7uzl3@bX0@zx@=E*O!;1{8zOav5xL;~CJ1h?qm(jJ2N zshxb0>m}e2Dg;QG7dx)84y)$9uJY9^mx<6pASD11-zvY|@8G+a7J7Xr7z$f`bj;%= zqtBMv%0G|=ieSX5SxU$_FA{s=fYW569iZ*!+3otmSS9p25Ki5>$>?PI>~}K9K%ZsU z((!NLG-s#)bY202EI#0|7QR636gV_rLvto|$d76Td4OxZp=HT;RYH-g;HD=9x*AJ* z7J~vhJKUBM(lm<^GIJ1u2K+!0M#iA?_-+5ZyJO}sd!E1#3+Kejg6suN$J%bYqJIsu zv5&!EYP8x?h=pHfCj7X{b66!%F1No`#3)~N=t@Y_y z7SJ@0>&>`{^<4YJZJqKa_p#*ZDL-tL(fUU?3^k!wQ7NvBTN4A|z#-Lu0Bi9TH;u&} z7GME%JcqmcsXlth=t+XFNe3Uj|x5TK(;p0lKVW=0_ zCz%qJc49+x53ar1qp(9>L*s=Aul_3M6PKFC#jIp0p~p*4(E=bRC?by@jFP8Bku>Nw zHstv>hRp5QA|MxxE>QwxY&Z{L#2Ls(VSxLmS4QZqwlcOd*?rc3x4PV@L`5%^P-1te he__G@U6ynPGJaEJX#n4OMY*j880nemR^M=n{V%{uvY7w? literal 0 HcmV?d00001 diff --git a/doc/LectureNotes/_build/html/_sources/chapter1.ipynb b/doc/LectureNotes/_build/html/_sources/chapter1.ipynb index 7e2ad50e8..4ec42071b 100644 --- a/doc/LectureNotes/_build/html/_sources/chapter1.ipynb +++ b/doc/LectureNotes/_build/html/_sources/chapter1.ipynb @@ -4,7 +4,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# Linear Regression, basic Elements\n", + "# Linear Regression\n", "\n", "[Video of Lecture](https://www.uio.no/studier/emner/matnat/fys/FYS-STK3155/h20/forelesningsvideoer/LectureAug21.mp4?vrtx=view-as-webpage)\n", "\n", @@ -15,7 +15,7 @@ "\n", "\n", "\n", - "Our emphasis throughout this series of lectures \n", + "Our emphasis throughout this series of lectures (small change) \n", "is on understanding the mathematical aspects of\n", "different algorithms used in the fields of data analysis and machine learning. \n", "\n", @@ -137,951 +137,98 @@ " * Clustering: Data are divided into groups with certain common traits, without knowing the different groups beforehand. It is thus a form of unsupervised learning.\n", "\n", "The methods we cover have three main topics in common, irrespective of\n", - "whether we deal with supervised or unsupervised learning. The first\n", - "ingredient is normally our data set (which can be subdivided into\n", - "training and test data), the second item is a model which is normally a\n", - "function of some parameters. The model reflects our knowledge of the system (or lack thereof). As an example, if we know that our data show a behavior similar to what would be predicted by a polynomial, fitting our data to a polynomial of some degree would then determin our model. \n", + "whether we deal with supervised or unsupervised learning.\n", + "* The first ingredient is normally our data set (which can be subdivided into training, validation and test data). Many find the most difficult part of using Machine Learning to be the set up of your data in a meaningful way. \n", "\n", - "The last ingredient is a so-called **cost**\n", - "function which allows us to present an estimate on how good our model\n", - "is in reproducing the data it is supposed to train. \n", - "At the heart of basically all ML algorithms there are so-called minimization algorithms, often we end up with various variants of **gradient** methods.\n", + "* The second item is a model which is normally a function of some parameters. The model reflects our knowledge of the system (or lack thereof). As an example, if we know that our data show a behavior similar to what would be predicted by a polynomial, fitting our data to a polynomial of some degree would then determin our model. \n", "\n", + "* The last ingredient is a so-called **cost/loss** function (or error or risk function) which allows us to present an estimate on how good our model is in reproducing the data it is supposed to train. \n", "\n", "\n", "\n", + "At the heart of basically all Machine Learning algorithms we will encounter so-called minimization or optimization algorithms. A large family of such methods are so-called **gradient methods**.\n", "\n", "\n", + "### A Frequentist approach to data analysis\n", "\n", - "## Software and needed installations\n", + "When you hear phrases like **predictions and estimations** and\n", + "**correlations and causations**, what do you think of? May be you think\n", + "of the difference between classifying new data points and generating\n", + "new data points.\n", + "Or perhaps you consider that correlations represent some kind of symmetric statements like\n", + "if $A$ is correlated with $B$, then $B$ is correlated with\n", + "$A$. Causation on the other hand is directional, that is if $A$ causes $B$, $B$ does not\n", + "necessarily cause $A$.\n", "\n", - "We will make extensive use of Python as programming language and its\n", - "myriad of available libraries. You will find\n", - "Jupyter notebooks invaluable in your work. You can run **R**\n", - "codes in the Jupyter/IPython notebooks, with the immediate benefit of\n", - "visualizing your data. You can also use compiled languages like C++,\n", - "Rust, Julia, Fortran etc if you prefer. The focus in these lectures will be\n", - "on Python.\n", + "These concepts are in some sense the difference between machine\n", + "learning and statistics. In machine learning and prediction based\n", + "tasks, we are often interested in developing algorithms that are\n", + "capable of learning patterns from given data in an automated fashion,\n", + "and then using these learned patterns to make predictions or\n", + "assessments of newly given data. In many cases, our primary concern\n", + "is the quality of the predictions or assessments, and we are less\n", + "concerned about the underlying patterns that were learned in order\n", + "to make these predictions.\n", "\n", + "In machine learning we normally use [a so-called frequentist approach](https://en.wikipedia.org/wiki/Frequentist_inference),\n", + "where the aim is to make predictions and find correlations. We focus\n", + "less on for example extracting a probability distribution function (PDF). The PDF can be\n", + "used in turn to make estimations and find causations such as given $A$\n", + "what is the likelihood of finding $B$.\n", "\n", - "If you have Python installed (we strongly recommend Python3) and you feel\n", - "pretty familiar with installing different packages, we recommend that\n", - "you install the following Python packages via **pip** as \n", "\n", - "1. pip install numpy scipy matplotlib ipython scikit-learn mglearn sympy pandas pillow \n", + "### What is a good model?\n", "\n", - "For Python3, replace **pip** with **pip3**.\n", + "In science and engineering we often end up in situations where we want to infer (or learn) a\n", + "quantitative model $M$ for a given set of sample points $\\boldsymbol{X} \\in [x_1, x_2,\\dots x_N]$.\n", "\n", - "For OSX users we recommend, after having installed Xcode, to\n", - "install **brew**. Brew allows for a seamless installation of additional\n", - "software via for example \n", + "As we will see repeatedely in these lectures, we could try to fit these data points to a model given by a\n", + "straight line, or if we wish to be more sophisticated to a more complex\n", + "function.\n", "\n", - "1. brew install python3\n", + "The reason for inferring such a model is that it\n", + "serves many useful purposes. On the one hand, the model can reveal information\n", + "encoded in the data or underlying mechanisms from which the data were generated. For instance, we could discover important\n", + "corelations that relate interesting physics interpretations.\n", "\n", - "For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution,\n", - "you can use **pip** as well and simply install Python as \n", + "In addition, it can simplify the representation of the given data set and help\n", + "us in making predictions about future data samples.\n", "\n", - "1. sudo apt-get install python3 (or python for pyhton2.7)\n", + "A first important consideration to keep in mind is that inferring the *correct* model\n", + "for a given data set is an elusive, if not impossible, task. The fundamental difficulty\n", + "is that if we are not specific about what we mean by a *correct* model, there\n", + "could easily be many different models that fit the given data set *equally well*.\n", "\n", - "etc etc. \n", "\n", "\n", "\n", - "## Python installers\n", + "The central question is this: what leads us to say that a model is correct or\n", + "optimal for a given data set? To make the model inference problem well posed, i.e.,\n", + "to guarantee that there is a unique optimal model for the given data, we need to\n", + "impose additional assumptions or restrictions on the class of models considered. To\n", + "this end, we should not be looking for just any model that can describe the data.\n", + "Instead, we should look for a **model** $M$ that is the best among a restricted class\n", + "of models. In addition, to make the model inference problem computationally\n", + "tractable, we need to specify how restricted the class of models needs to be. A\n", + "common strategy is to start \n", + "with the simplest possible class of models that is just necessary to describe the data\n", + "or solve the problem at hand. More precisely, the model class should be rich enough\n", + "to contain at least one model that can fit the data to a desired accuracy and yet be\n", + "restricted enough that it is relatively simple to find the best model for the given data.\n", "\n", - "If you don't want to perform these operations separately and venture\n", - "into the hassle of exploring how to set up dependencies and paths, we\n", - "recommend two widely used distrubutions which set up all relevant\n", - "dependencies for Python, namely \n", + "Thus, the most popular strategy is to start from the\n", + "simplest class of models and increase the complexity of the models only when the\n", + "simpler models become inadequate. For instance, if we work with a regression problem to fit a set of sample points, one\n", + "may first try the simplest class of models, namely linear models, followed obviously by more complex models.\n", "\n", - "* [Anaconda](https://docs.anaconda.com/), \n", + "How to evaluate which model fits best the data is something we will come back to over and over again in these sets of lectures.\n", "\n", - "which is an open source\n", - "distribution of the Python and R programming languages for large-scale\n", - "data processing, predictive analytics, and scientific computing, that\n", - "aims to simplify package management and deployment. Package versions\n", - "are managed by the package management system **conda**. \n", "\n", - "* [Enthought canopy](https://www.enthought.com/product/canopy/) \n", "\n", - "is a Python\n", - "distribution for scientific and analytic computing distribution and\n", - "analysis environment, available for free and under a commercial\n", - "license.\n", "\n", - "Furthermore, [Google's Colab](https://colab.research.google.com/notebooks/welcome.ipynb) is a free Jupyter notebook environment that requires \n", - "no setup and runs entirely in the cloud. Try it out!\n", "\n", "\n", - "## Useful Python libraries\n", - "Here we list several useful Python libraries we strongly recommend (if you use anaconda many of these are already there)\n", "\n", - "* [NumPy](https://www.numpy.org/) is a highly popular library for large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays\n", - "\n", - "* [The pandas](https://pandas.pydata.org/) library provides high-performance, easy-to-use data structures and data analysis tools \n", - "\n", - "* [Xarray](http://xarray.pydata.org/en/stable/) is a Python package that makes working with labelled multi-dimensional arrays simple, efficient, and fun!\n", - "\n", - "* [Scipy](https://www.scipy.org/) (pronounced “Sigh Pie”) is a Python-based ecosystem of open-source software for mathematics, science, and engineering. \n", - "\n", - "* [Matplotlib](https://matplotlib.org/) is a Python 2D plotting library which produces publication quality figures in a variety of hardcopy formats and interactive environments across platforms.\n", - "\n", - "* [Autograd](https://github.com/HIPS/autograd) can automatically differentiate native Python and Numpy code. It can handle a large subset of Python's features, including loops, ifs, recursion and closures, and it can even take derivatives of derivatives of derivatives\n", - "\n", - "* [SymPy](https://www.sympy.org/en/index.html) is a Python library for symbolic mathematics. \n", - "\n", - "* [scikit-learn](https://scikit-learn.org/stable/) has simple and efficient tools for machine learning, data mining and data analysis\n", - "\n", - "* [TensorFlow](https://www.tensorflow.org/) is a Python library for fast numerical computing created and released by Google\n", - "\n", - "* [Keras](https://keras.io/) is a high-level neural networks API, written in Python and capable of running on top of TensorFlow, CNTK, or Theano\n", - "\n", - "* And many more such as [pytorch](https://pytorch.org/), [Theano](https://pypi.org/project/Theano/) etc \n", - "\n", - "## Installing R, C++, cython or Julia\n", - "\n", - "You will also find it convenient to utilize **R**. We will mainly\n", - "use Python during our lectures and in various projects and exercises.\n", - "Those of you\n", - "already familiar with **R** should feel free to continue using **R**, keeping\n", - "however an eye on the parallel Python set ups. Similarly, if you are a\n", - "Python afecionado, feel free to explore **R** as well. Jupyter/Ipython\n", - "notebook allows you to run **R** codes interactively in your\n", - "browser. The software library **R** is really tailored for statistical data analysis\n", - "and allows for an easy usage of the tools and algorithms we will discuss in these\n", - "lectures.\n", - "\n", - "To install **R** with Jupyter notebook \n", - "[follow the link here](https://mpacer.org/maths/r-kernel-for-ipython-notebook)\n", - "\n", - "\n", - "\n", - "\n", - "## Installing R, C++, cython, Numba etc\n", - "\n", - "\n", - "For the C++ aficionados, Jupyter/IPython notebook allows you also to\n", - "install C++ and run codes written in this language interactively in\n", - "the browser. Since we will emphasize writing many of the algorithms\n", - "yourself, you can thus opt for either Python or C++ (or Fortran or other compiled languages) as programming\n", - "languages.\n", - "\n", - "To add more entropy, **cython** can also be used when running your\n", - "notebooks. It means that Python with the jupyter notebook\n", - "setup allows you to integrate widely popular softwares and tools for\n", - "scientific computing. Similarly, the \n", - "[Numba Python package](https://numba.pydata.org/) delivers increased performance\n", - "capabilities with minimal rewrites of your codes. With its\n", - "versatility, including symbolic operations, Python offers a unique\n", - "computational environment. Your jupyter notebook can easily be\n", - "converted into a nicely rendered **PDF** file or a Latex file for\n", - "further processing. For example, convert to latex as" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - " pycod jupyter nbconvert filename.ipynb --to latex \n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "And to add more versatility, the Python package [SymPy](http://www.sympy.org/en/index.html) is a Python library for symbolic mathematics. It aims to become a full-featured computer algebra system (CAS) and is entirely written in Python. \n", - "\n", - "Finally, if you wish to use the light mark-up language \n", - "[doconce](https://github.com/hplgit/doconce) you can convert a standard ascii text file into various HTML \n", - "formats, ipython notebooks, latex files, pdf files etc with minimal edits. These lectures were generated using **doconce**.\n", - "\n", - "\n", - "\n", - "## Numpy examples and Important Matrix and vector handling packages\n", - "\n", - "There are several central software libraries for linear algebra and eigenvalue problems. Several of the more\n", - "popular ones have been wrapped into ofter software packages like those from the widely used text **Numerical Recipes**. The original source codes in many of the available packages are often taken from the widely used\n", - "software package LAPACK, which follows two other popular packages\n", - "developed in the 1970s, namely EISPACK and LINPACK. We describe them shortly here.\n", - "\n", - " * LINPACK: package for linear equations and least square problems.\n", - "\n", - " * LAPACK:package for solving symmetric, unsymmetric and generalized eigenvalue problems. From LAPACK's website it is possible to download for free all source codes from this library. Both C/C++ and Fortran versions are available.\n", - "\n", - " * BLAS (I, II and III): (Basic Linear Algebra Subprograms) are routines that provide standard building blocks for performing basic vector and matrix operations. Blas I is vector operations, II vector-matrix operations and III matrix-matrix operations. Highly parallelized and efficient codes, all available for download from .\n", - "\n", - "## Basic Matrix Features\n", - "\n", - "Matrix properties reminder" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\mathbf{A} =\n", - " \\begin{bmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\\\\n", - " a_{21} & a_{22} & a_{23} & a_{24} \\\\\n", - " a_{31} & a_{32} & a_{33} & a_{34} \\\\\n", - " a_{41} & a_{42} & a_{43} & a_{44}\n", - " \\end{bmatrix}\\qquad\n", - "\\mathbf{I} =\n", - " \\begin{bmatrix} 1 & 0 & 0 & 0 \\\\\n", - " 0 & 1 & 0 & 0 \\\\\n", - " 0 & 0 & 1 & 0 \\\\\n", - " 0 & 0 & 0 & 1\n", - " \\end{bmatrix}\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The inverse of a matrix is defined by" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\mathbf{A}^{-1} \\cdot \\mathbf{A} = I\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "
    Relations Name matrix elements
    $A = A^{T}$ symmetric $a_{ij} = a_{ji}$
    $A = \\left (A^{T} \\right )^{-1}$ real orthogonal $\\sum_k a_{ik} a_{jk} = \\sum_k a_{ki} a_{kj} = \\delta_{ij}$
    $A = A^{ * }$ real matrix $a_{ij} = a_{ij}^{ * }$
    $A = A^{\\dagger}$ hermitian $a_{ij} = a_{ji}^{ * }$
    $A = \\left (A^{\\dagger} \\right )^{-1}$ unitary $\\sum_k a_{ik} a_{jk}^{ * } = \\sum_k a_{ki}^{ * } a_{kj} = \\delta_{ij}$
    \n", - "\n", - "\n", - "### Some famous Matrices\n", - "\n", - " * Diagonal if $a_{ij}=0$ for $i\\ne j$\n", - "\n", - " * Upper triangular if $a_{ij}=0$ for $i > j$\n", - "\n", - " * Lower triangular if $a_{ij}=0$ for $i < j$\n", - "\n", - " * Upper Hessenberg if $a_{ij}=0$ for $i > j+1$\n", - "\n", - " * Lower Hessenberg if $a_{ij}=0$ for $i < j+1$\n", - "\n", - " * Tridiagonal if $a_{ij}=0$ for $|i -j| > 1$\n", - "\n", - " * Lower banded with bandwidth $p$: $a_{ij}=0$ for $i > j+p$\n", - "\n", - " * Upper banded with bandwidth $p$: $a_{ij}=0$ for $i < j+p$\n", - "\n", - " * Banded, block upper triangular, block lower triangular....\n", - "\n", - "### More Basic Matrix Features\n", - "\n", - "Some Equivalent Statements\n", - "For an $N\\times N$ matrix $\\mathbf{A}$ the following properties are all equivalent\n", - "\n", - " * If the inverse of $\\mathbf{A}$ exists, $\\mathbf{A}$ is nonsingular.\n", - "\n", - " * The equation $\\mathbf{Ax}=0$ implies $\\mathbf{x}=0$.\n", - "\n", - " * The rows of $\\mathbf{A}$ form a basis of $R^N$.\n", - "\n", - " * The columns of $\\mathbf{A}$ form a basis of $R^N$.\n", - "\n", - " * $\\mathbf{A}$ is a product of elementary matrices.\n", - "\n", - " * $0$ is not eigenvalue of $\\mathbf{A}$.\n", - "\n", - "## Numpy and arrays\n", - "[Numpy](http://www.numpy.org/) provides an easy way to handle arrays in Python. The standard way to import this library is as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Here follows a simple example where we set up an array of ten elements, all determined by random numbers drawn according to the normal distribution," - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "n = 10\n", - "x = np.random.normal(size=n)\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We defined a vector $x$ with $n=10$ elements with its values given by the Normal distribution $N(0,1)$.\n", - "Another alternative is to declare a vector as follows" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "x = np.array([1, 2, 3])\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Here we have defined a vector with three elements, with $x_0=1$, $x_1=2$ and $x_2=3$. Note that both Python and C++\n", - "start numbering array elements from $0$ and on. This means that a vector with $n$ elements has a sequence of entities $x_0, x_1, x_2, \\dots, x_{n-1}$. We could also let (recommended) Numpy to compute the logarithms of a specific array as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "x = np.log(np.array([4, 7, 8]))\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In the last example we used Numpy's unary function $np.log$. This function is\n", - "highly tuned to compute array elements since the code is vectorized\n", - "and does not require looping. We normaly recommend that you use the\n", - "Numpy intrinsic functions instead of the corresponding **log** function\n", - "from Python's **math** module. The looping is done explicitely by the\n", - "**np.log** function. The alternative, and slower way to compute the\n", - "logarithms of a vector would be to write" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "from math import log\n", - "x = np.array([4, 7, 8])\n", - "for i in range(0, len(x)):\n", - " x[i] = log(x[i])\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We note that our code is much longer already and we need to import the **log** function from the **math** module. \n", - "The attentive reader will also notice that the output is $[1, 1, 2]$. Python interprets automagically our numbers as integers (like the **automatic** keyword in C++). To change this we could define our array elements to be double precision numbers as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "x = np.log(np.array([4, 7, 8], dtype = np.float64))\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "or simply write them as double precision numbers (Python uses 64 bits as default for floating point type variables), that is" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "x = np.log(np.array([4.0, 7.0, 8.0])\n", - "print(x)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "To check the number of bytes (remember that one byte contains eight bits for double precision variables), you can use simple use the **itemsize** functionality (the array $x$ is actually an object which inherits the functionalities defined in Numpy) as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "x = np.log(np.array([4.0, 7.0, 8.0])\n", - "print(x.itemsize)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Matrices in Python\n", - "\n", - "Having defined vectors, we are now ready to try out matrices. We can\n", - "define a $3 \\times 3 $ real matrix $\\hat{A}$ as (recall that we user\n", - "lowercase letters for vectors and uppercase letters for matrices)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))\n", - "print(A)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "If we use the **shape** function we would get $(3, 3)$ as output, that is verifying that our matrix is a $3\\times 3$ matrix. We can slice the matrix and print for example the first column (Python organized matrix elements in a row-major order, see below) as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))\n", - "# print the first column, row-major order and elements start with 0\n", - "print(A[:,0])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can continue this was by printing out other columns or rows. The example here prints out the second column" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))\n", - "# print the first column, row-major order and elements start with 0\n", - "print(A[1,:])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Numpy contains many other functionalities that allow us to slice, subdivide etc etc arrays. We strongly recommend that you look up the [Numpy website for more details](http://www.numpy.org/). Useful functions when defining a matrix are the **np.zeros** function which declares a matrix of a given dimension and sets all elements to zero" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "n = 10\n", - "# define a matrix of dimension 10 x 10 and set all elements to zero\n", - "A = np.zeros( (n, n) )\n", - "print(A)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "or initializing all elements to" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "n = 10\n", - "# define a matrix of dimension 10 x 10 and set all elements to one\n", - "A = np.ones( (n, n) )\n", - "print(A)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "or as unitarily distributed random numbers (see the material on random number generators in the statistics part)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "n = 10\n", - "# define a matrix of dimension 10 x 10 and set all elements to random numbers with x \\in [0, 1]\n", - "A = np.random.rand(n, n)\n", - "print(A)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "As we will see throughout these lectures, there are several extremely useful functionalities in Numpy.\n", - "As an example, consider the discussion of the covariance matrix. Suppose we have defined three vectors\n", - "$\\hat{x}, \\hat{y}, \\hat{z}$ with $n$ elements each. The covariance matrix is defined as" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\hat{\\Sigma} = \\begin{bmatrix} \\sigma_{xx} & \\sigma_{xy} & \\sigma_{xz} \\\\\n", - " \\sigma_{yx} & \\sigma_{yy} & \\sigma_{yz} \\\\\n", - " \\sigma_{zx} & \\sigma_{zy} & \\sigma_{zz} \n", - " \\end{bmatrix},\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "where for example" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\sigma_{xy} =\\frac{1}{n} \\sum_{i=0}^{n-1}(x_i- \\overline{x})(y_i- \\overline{y}).\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The Numpy function **np.cov** calculates the covariance elements using the factor $1/(n-1)$ instead of $1/n$ since it assumes we do not have the exact mean values. \n", - "The following simple function uses the **np.vstack** function which takes each vector of dimension $1\\times n$ and produces a $3\\times n$ matrix $\\hat{W}$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\hat{W} = \\begin{bmatrix} x_0 & y_0 & z_0 \\\\\n", - " x_1 & y_1 & z_1 \\\\\n", - " x_2 & y_2 & z_2 \\\\\n", - " \\dots & \\dots & \\dots \\\\\n", - " x_{n-2} & y_{n-2} & z_{n-2} \\\\\n", - " x_{n-1} & y_{n-1} & z_{n-1}\n", - " \\end{bmatrix},\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "which in turn is converted into into the $3\\times 3$ covariance matrix\n", - "$\\hat{\\Sigma}$ via the Numpy function **np.cov()**. We note that we can also calculate\n", - "the mean value of each set of samples $\\hat{x}$ etc using the Numpy\n", - "function **np.mean(x)**. We can also extract the eigenvalues of the\n", - "covariance matrix through the **np.linalg.eig()** function." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "# Importing various packages\n", - "import numpy as np\n", - "\n", - "n = 100\n", - "x = np.random.normal(size=n)\n", - "print(np.mean(x))\n", - "y = 4+3*x+np.random.normal(size=n)\n", - "print(np.mean(y))\n", - "z = x**3+np.random.normal(size=n)\n", - "print(np.mean(z))\n", - "W = np.vstack((x, y, z))\n", - "Sigma = np.cov(W)\n", - "print(Sigma)\n", - "Eigvals, Eigvecs = np.linalg.eig(Sigma)\n", - "print(Eigvals)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "%matplotlib inline\n", - "\n", - "import numpy as np\n", - "import matplotlib.pyplot as plt\n", - "from scipy import sparse\n", - "eye = np.eye(4)\n", - "print(eye)\n", - "sparse_mtx = sparse.csr_matrix(eye)\n", - "print(sparse_mtx)\n", - "x = np.linspace(-10,10,100)\n", - "y = np.sin(x)\n", - "plt.plot(x,y,marker='x')\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Meet the Pandas\n", - "\n", - "\n", - "\n", - "\n", - "Another useful Python package is\n", - "[pandas](https://pandas.pydata.org/), which is an open source library\n", - "providing high-performance, easy-to-use data structures and data\n", - "analysis tools for Python. **pandas** stands for panel data, a term borrowed from econometrics and is an efficient library for data analysis with an emphasis on tabular data.\n", - "**pandas** has two major classes, the **DataFrame** class with two-dimensional data objects and tabular data organized in columns and the class **Series** with a focus on one-dimensional data objects. Both classes allow you to index data easily as we will see in the examples below. \n", - "**pandas** allows you also to perform mathematical operations on the data, spanning from simple reshapings of vectors and matrices to statistical operations. \n", - "\n", - "The following simple example shows how we can, in an easy way make tables of our data. Here we define a data set which includes names, place of birth and date of birth, and displays the data in an easy to read way. We will see repeated use of **pandas**, in particular in connection with classification of data." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import pandas as pd\n", - "from IPython.display import display\n", - "data = {'First Name': [\"Frodo\", \"Bilbo\", \"Aragorn II\", \"Samwise\"],\n", - " 'Last Name': [\"Baggins\", \"Baggins\",\"Elessar\",\"Gamgee\"],\n", - " 'Place of birth': [\"Shire\", \"Shire\", \"Eriador\", \"Shire\"],\n", - " 'Date of Birth T.A.': [2968, 2890, 2931, 2980]\n", - " }\n", - "data_pandas = pd.DataFrame(data)\n", - "display(data_pandas)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In the above we have imported **pandas** with the shorthand **pd**, the latter has become the standard way we import **pandas**. We make then a list of various variables\n", - "and reorganize the aboves lists into a **DataFrame** and then print out a neat table with specific column labels as *Name*, *place of birth* and *date of birth*.\n", - "Displaying these results, we see that the indices are given by the default numbers from zero to three.\n", - "**pandas** is extremely flexible and we can easily change the above indices by defining a new type of indexing as" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "data_pandas = pd.DataFrame(data,index=['Frodo','Bilbo','Aragorn','Sam'])\n", - "display(data_pandas)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Thereafter we display the content of the row which begins with the index **Aragorn**" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "display(data_pandas.loc['Aragorn'])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can easily append data to this, for example" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "new_hobbit = {'First Name': [\"Peregrin\"],\n", - " 'Last Name': [\"Took\"],\n", - " 'Place of birth': [\"Shire\"],\n", - " 'Date of Birth T.A.': [2990]\n", - " }\n", - "data_pandas=data_pandas.append(pd.DataFrame(new_hobbit, index=['Pippin']))\n", - "display(data_pandas)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Here are other examples where we use the **DataFrame** functionality to handle arrays, now with more interesting features for us, namely numbers. We set up a matrix \n", - "of dimensionality $10\\times 5$ and compute the mean value and standard deviation of each column. Similarly, we can perform mathematial operations like squaring the matrix elements and many other operations." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "import numpy as np\n", - "import pandas as pd\n", - "from IPython.display import display\n", - "np.random.seed(100)\n", - "# setting up a 10 x 5 matrix\n", - "rows = 10\n", - "cols = 5\n", - "a = np.random.randn(rows,cols)\n", - "df = pd.DataFrame(a)\n", - "display(df)\n", - "print(df.mean())\n", - "print(df.std())\n", - "display(df**2)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Thereafter we can select specific columns only and plot final results" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "df.columns = ['First', 'Second', 'Third', 'Fourth', 'Fifth']\n", - "df.index = np.arange(10)\n", - "\n", - "display(df)\n", - "print(df['Second'].mean() )\n", - "print(df.info())\n", - "print(df.describe())\n", - "\n", - "from pylab import plt, mpl\n", - "plt.style.use('seaborn')\n", - "mpl.rcParams['font.family'] = 'serif'\n", - "\n", - "df.cumsum().plot(lw=2.0, figsize=(10,6))\n", - "plt.show()\n", - "\n", - "\n", - "df.plot.bar(figsize=(10,6), rot=15)\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can produce a $4\\times 4$ matrix" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "editable": true - }, - "outputs": [], - "source": [ - "b = np.arange(16).reshape((4,4))\n", - "print(b)\n", - "df1 = pd.DataFrame(b)\n", - "print(df1)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "and many other operations. \n", - "\n", - "The **Series** class is another important class included in\n", - "**pandas**. You can view it as a specialization of **DataFrame** but where\n", - "we have just a single column of data. It shares many of the same features as _DataFrame. As with **DataFrame**,\n", - "most operations are vectorized, achieving thereby a high performance when dealing with computations of arrays, in particular labeled arrays.\n", - "As we will see below it leads also to a very concice code close to the mathematical operations we may be interested in.\n", - "For multidimensional arrays, we recommend strongly [xarray](http://xarray.pydata.org/en/stable/). **xarray** has much of the same flexibility as **pandas**, but allows for the extension to higher dimensions than two. We will see examples later of the usage of both **pandas** and **xarray**. \n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "In order to study various Machine Learning algorithms, we need to\n", - "access data. Acccessing data is an essential step in all machine\n", - "learning algorithms. In particular, setting up the so-called **design\n", - "matrix** (to be defined below) is often the first element we need in\n", - "order to perform our calculations. To set up the design matrix means\n", - "reading (and later, when the calculations are done, writing) data\n", - "in various formats, The formats span from reading files from disk,\n", - "loading data from databases and interacting with online sources\n", - "like web application programming interfaces (APIs).\n", - "\n", - "In handling various input formats, as discussed above, we will mainly stay with **pandas**,\n", - "a Python package which allows us, in a seamless and painless way, to\n", - "deal with a multitude of formats, from standard **csv** (comma separated\n", - "values) files, via **excel**, **html** to **hdf5** formats. With **pandas**\n", - "and the **DataFrame** and **Series** functionalities we are able to convert text data\n", - "into the calculational formats we need for a specific algorithm. And our code is going to be \n", - "pretty close the basic mathematical expressions.\n", - "\n", - "Our first data set is going to be a classic from nuclear physics, namely all\n", - "available data on binding energies. Don't be intimidated if you are not familiar with nuclear physics. It serves simply as an example here of a data set. \n", - "\n", - "We will show some of the\n", - "strengths of packages like **Scikit-Learn** in fitting nuclear binding energies to\n", - "specific functions using linear regression first. Then, as a teaser, we will show you how \n", - "you can easily implement other algorithms like decision trees and random forests and neural networks.\n", - "\n", - "But before we really start with nuclear physics data, let's just look at some simpler polynomial fitting cases, such as,\n", - "(don't be offended) fitting straight lines!\n", "\n", "\n", "\n", @@ -1153,6 +300,8 @@ }, "outputs": [], "source": [ + "%matplotlib inline\n", + "\n", "# Importing various packages\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", @@ -1480,7 +629,8 @@ "metadata": {}, "source": [ "$$\n", - "H_{\\delta}(a)={\\begin{cases}{\\frac {1}{2}}{a^{2}}&{\\text{for }}|a|\\leq \\delta ,\\\\\\delta (|a|-{\\frac {1}{2}}\\delta ),&{\\text{otherwise.}}\\end{cases}}}.\n", + "H_{\\delta}(a)=\\left\\{\\begin{array}\\frac{1}{2}a^{2}&{\\text{for }}|a|\\leq \\delta ,\\\\ \\delta (|a|-\\frac {1}{2}\\del\\\n", + "ta ),&{\\text{otherwise.}\\end{array}\\right.\n", "$$" ] }, @@ -2741,7 +1891,7 @@ "metadata": {}, "source": [ "4\n", - "8\n", + "3\n", " \n", "<\n", "<\n", @@ -2765,7 +1915,7 @@ "metadata": {}, "source": [ "4\n", - "9\n", + "4\n", " \n", "<\n", "<\n", @@ -2788,8 +1938,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "4\n", "5\n", - "0\n", " \n", "<\n", "<\n", @@ -4058,6 +3208,348 @@ "print(\"MSE after scaling: {:.2f}\".format(mean_squared_error(clf.predict(X_test_scaled), y_test)))\n", "print(\"R2 score for scaled data: {:.2f}\".format(clf.score(X_test_scaled,y_test)))" ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Exercises\n", + "\n", + "### Exercise: Setting up various Python environments\n", + "\n", + "The first exercise here is of a mere technical art. We want you to have \n", + "* git as a version control software and to establish a user account on a provider like GitHub. Other providers like GitLab etc are equally fine. You can also use the University of Oslo [GitHub facilities](https://www.uio.no/tjenester/it/maskin/filer/versjonskontroll/github.html). \n", + "\n", + "* Install various Python packages\n", + "\n", + "We will make extensive use of Python as programming language and its\n", + "myriad of available libraries. You will find\n", + "IPython/Jupyter notebooks invaluable in your work. You can run **R**\n", + "codes in the Jupyter/IPython notebooks, with the immediate benefit of\n", + "visualizing your data. You can also use compiled languages like C++,\n", + "Rust, Fortran etc if you prefer. The focus in these lectures will be\n", + "on Python.\n", + "\n", + "If you have Python installed (we recommend Python3) and you feel\n", + "pretty familiar with installing different packages, we recommend that\n", + "you install the following Python packages via **pip** as \n", + "\n", + "1. pip install numpy scipy matplotlib ipython scikit-learn sympy pandas pillow \n", + "\n", + "For **Tensorflow**, we recommend following the instructions in the text of \n", + "[Aurelien Geron, Hands‑On Machine Learning with Scikit‑Learn and TensorFlow, O'Reilly](http://shop.oreilly.com/product/0636920052289.do)\n", + "\n", + "We will come back to **tensorflow** later. \n", + "\n", + "For Python3, replace **pip** with **pip3**.\n", + "\n", + "For OSX users we recommend, after having installed Xcode, to\n", + "install **brew**. Brew allows for a seamless installation of additional\n", + "software via for example \n", + "\n", + "1. brew install python3\n", + "\n", + "For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution,\n", + "you can use **pip** as well and simply install Python as \n", + "\n", + "1. sudo apt-get install python3 (or python for Python2.7)\n", + "\n", + "If you don't want to perform these operations separately and venture\n", + "into the hassle of exploring how to set up dependencies and paths, we\n", + "recommend two widely used distrubutions which set up all relevant\n", + "dependencies for Python, namely \n", + "\n", + "* [Anaconda](https://docs.anaconda.com/), \n", + "\n", + "which is an open source\n", + "distribution of the Python and R programming languages for large-scale\n", + "data processing, predictive analytics, and scientific computing, that\n", + "aims to simplify package management and deployment. Package versions\n", + "are managed by the package management system **conda**. \n", + "\n", + "* [Enthought canopy](https://www.enthought.com/product/canopy/) \n", + "\n", + "is a Python\n", + "distribution for scientific and analytic computing distribution and\n", + "analysis environment, available for free and under a commercial\n", + "license.\n", + "\n", + "We recommend using **Anaconda** if you are not too familiar with setting paths in a terminal environment.\n", + "\n", + "\n", + "\n", + "\n", + "### Exercise: making your own data and exploring scikit-learn\n", + "\n", + "We will generate our own dataset for a function $y(x)$ where $x \\in [0,1]$ and defined by random numbers computed with the uniform distribution. The function $y$ is a quadratic polynomial in $x$ with added stochastic noise according to the normal distribution $\\cal {N}(0,1)$.\n", + "The following simple Python instructions define our $x$ and $y$ values (with 100 data points)." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "x = np.random.rand(100,1)\n", + "y = 2.0+5*x*x+0.1*np.random.randn(100,1)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "1. Write your own code (following the examples under the [regression notes](https://compphysics.github.io/MachineLearning/doc/LectureNotes/_build/html/chapter1.html)) for computing the parametrization of the data set fitting a second-order polynomial. \n", + "\n", + "2. Use thereafter **scikit-learn** (see again the examples in the regression slides) and compare with your own code. \n", + "\n", + "3. Using scikit-learn, compute also the mean square error, a risk metric corresponding to the expected value of the squared (quadratic) error defined as" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "$$\n", + "MSE(\\hat{y},\\hat{\\tilde{y}}) = \\frac{1}{n}\n", + "\\sum_{i=0}^{n-1}(y_i-\\tilde{y}_i)^2,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "and the $R^2$ score function.\n", + "If $\\tilde{\\hat{y}}_i$ is the predicted value of the $i-th$ sample and $y_i$ is the corresponding true value, then the score $R^2$ is defined as" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "$$\n", + "R^2(\\hat{y}, \\tilde{\\hat{y}}) = 1 - \\frac{\\sum_{i=0}^{n - 1} (y_i - \\tilde{y}_i)^2}{\\sum_{i=0}^{n - 1} (y_i - \\bar{y})^2},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "where we have defined the mean value of $\\hat{y}$ as" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "$$\n", + "\\bar{y} = \\frac{1}{n} \\sum_{i=0}^{n - 1} y_i.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "You can use the functionality included in scikit-learn. If you feel for it, you can use your own program and define functions which compute the above two functions. \n", + "Discuss the meaning of these results. Try also to vary the coefficient in front of the added stochastic noise term and discuss the quality of the fits.\n", + "\n", + "!bsol\n", + "The code here is an example of where we define our own design matrix and fit parameters $\\beta$." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "import os\n", + "import numpy as np\n", + "import pandas as pd\n", + "import matplotlib.pyplot as plt\n", + "from sklearn.model_selection import train_test_split\n", + "\n", + "def save_fig(fig_id):\n", + " plt.savefig(image_path(fig_id) + \".png\", format='png')\n", + "\n", + "def R2(y_data, y_model):\n", + " return 1 - np.sum((y_data - y_model) ** 2) / np.sum((y_data - np.mean(y_data)) ** 2)\n", + "def MSE(y_data,y_model):\n", + " n = np.size(y_model)\n", + " return np.sum((y_data-y_model)**2)/n\n", + "\n", + "x = np.random.rand(100)\n", + "y = 2.0+5*x*x+0.1*np.random.randn(100)\n", + "\n", + "\n", + "# The design matrix now as function of a given polynomial\n", + "X = np.zeros((len(x),3))\n", + "X[:,0] = 1.0\n", + "X[:,1] = x\n", + "X[:,2] = x**2\n", + "# We split the data in test and training data\n", + "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n", + "# matrix inversion to find beta\n", + "beta = np.linalg.inv(X_train.T @ X_train) @ X_train.T @ y_train\n", + "print(beta)\n", + "# and then make the prediction\n", + "ytilde = X_train @ beta\n", + "print(\"Training R2\")\n", + "print(R2(y_train,ytilde))\n", + "print(\"Training MSE\")\n", + "print(MSE(y_train,ytilde))\n", + "ypredict = X_test @ beta\n", + "print(\"Test R2\")\n", + "print(R2(y_test,ypredict))\n", + "print(\"Test MSE\")\n", + "print(MSE(y_test,ypredict))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "!esol\n", + "\n", + "\n", + "\n", + "### Exercise: Normalizing our data\n", + "\n", + "A much used approach before starting to train the data is to preprocess our\n", + "data. Normally the data may need a rescaling and/or may be sensitive\n", + "to extreme values. Scaling the data renders our inputs much more\n", + "suitable for the algorithms we want to employ.\n", + "\n", + "**Scikit-Learn** has several functions which allow us to rescale the\n", + "data, normally resulting in much better results in terms of various\n", + "accuracy scores. The **StandardScaler** function in **Scikit-Learn**\n", + "ensures that for each feature/predictor we study the mean value is\n", + "zero and the variance is one (every column in the design/feature\n", + "matrix). This scaling has the drawback that it does not ensure that\n", + "we have a particular maximum or minimum in our data set. Another\n", + "function included in **Scikit-Learn** is the **MinMaxScaler** which\n", + "ensures that all features are exactly between $0$ and $1$. The\n", + "\n", + "\n", + "The **Normalizer** scales each data\n", + "point such that the feature vector has a euclidean length of one. In other words, it\n", + "projects a data point on the circle (or sphere in the case of higher dimensions) with a\n", + "radius of 1. This means every data point is scaled by a different number (by the\n", + "inverse of it’s length).\n", + "This normalization is often used when only the direction (or angle) of the data matters,\n", + "not the length of the feature vector.\n", + "\n", + "The **RobustScaler** works similarly to the StandardScaler in that it\n", + "ensures statistical properties for each feature that guarantee that\n", + "they are on the same scale. However, the RobustScaler uses the median\n", + "and quartiles, instead of mean and variance. This makes the\n", + "RobustScaler ignore data points that are very different from the rest\n", + "(like measurement errors). These odd data points are also called\n", + "outliers, and might often lead to trouble for other scaling\n", + "techniques.\n", + "\n", + "\n", + "It also common to split the data in a **training** set and a **testing** set. A typical split is to use $80\\%$ of the data for training and the rest\n", + "for testing. This can be done as follows with our design matrix $\\boldsymbol{X}$ and data $\\boldsymbol{y}$ (remember to import **scikit-learn**)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "# split in training and test data\n", + "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Then we can use the standard scaler to scale our data as" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "scaler = StandardScaler()\n", + "scaler.fit(X_train)\n", + "X_train_scaled = scaler.transform(X_train)\n", + "X_test_scaled = scaler.transform(X_test)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "In this exercise we want you to to compute the MSE for the training\n", + "data and the test data as function of the complexity of a polynomial,\n", + "that is the degree of a given polynomial. We want you also to compute the $R2$ score as function of the complexity of the model for both training data and test data. You should also run the calculation with and without scaling. \n", + "\n", + "One of \n", + "the aims is to reproduce Figure 2.11 of [Hastie et al](https://github.com/CompPhysics/MLErasmus/blob/master/doc/Textbooks/elementsstat.pdf).\n", + "\n", + "\n", + "\n", + "Our data is defined by $x\\in [-3,3]$ with a total of for example $100$ data points." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "np.random.seed()\n", + "n = 100\n", + "maxdegree = 14\n", + "# Make data set.\n", + "x = np.linspace(-3, 3, n).reshape(-1, 1)\n", + "y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "where $y$ is the function we want to fit with a given polynomial.\n", + "!bsubex\n", + "Write a first code which sets up a design matrix $X$ defined by a fifth-order polynomial. Scale your data and split it in training and test data. \n", + "!esubex\n", + "\n", + "!bsubex\n", + "Perform an ordinary least squares and compute the means squared error and the $R2$ factor for the training data and the test data, with and without scaling.\n", + "!esubex\n", + "\n", + "!bsubex\n", + "Add now a model which allows you to make polynomials up to degree $15$. Perform a standard OLS fitting of the training data and compute the MSE and $R2$ for the training and test data and plot both test and training data MSE and $R2$ as functions of the polynomial degree. Compare what you see with Figure 2.11 of Hastie et al. Comment your results. For which polynomial degree do you find an optimal MSE (smallest value)?\n", + "!bsol\n", + "Here you simply need to change the degree of the polynomial in the above code to $n=15$.\n", + "!esol\n", + "!esubex" + ] } ], "metadata": {}, diff --git a/doc/LectureNotes/_build/html/chapter1.html b/doc/LectureNotes/_build/html/chapter1.html index bdc472e82..e1802b5f1 100644 --- a/doc/LectureNotes/_build/html/chapter1.html +++ b/doc/LectureNotes/_build/html/chapter1.html @@ -5,7 +5,7 @@ - 3. Linear Regression, basic Elements — Applied Data Analysis and Machine Learning + 3. Linear Regression — Applied Data Analysis and Machine Learning @@ -139,7 +139,7 @@

    The methods we cover have three main topics in common, irrespective of -whether we deal with supervised or unsupervised learning. The first -ingredient is normally our data set (which can be subdivided into -training and test data), the second item is a model which is normally a -function of some parameters. The model reflects our knowledge of the system (or lack thereof). As an example, if we know that our data show a behavior similar to what would be predicted by a polynomial, fitting our data to a polynomial of some degree would then determin our model.

    -

    The last ingredient is a so-called cost -function which allows us to present an estimate on how good our model -is in reproducing the data it is supposed to train.
    -At the heart of basically all ML algorithms there are so-called minimization algorithms, often we end up with various variants of gradient methods.

    - -
    -

    3.3. Software and needed installations

    -

    We will make extensive use of Python as programming language and its -myriad of available libraries. You will find -Jupyter notebooks invaluable in your work. You can run R -codes in the Jupyter/IPython notebooks, with the immediate benefit of -visualizing your data. You can also use compiled languages like C++, -Rust, Julia, Fortran etc if you prefer. The focus in these lectures will be -on Python.

    -

    If you have Python installed (we strongly recommend Python3) and you feel -pretty familiar with installing different packages, we recommend that -you install the following Python packages via pip as

    -
      -

    1. pip install numpy scipy matplotlib ipython scikit-learn mglearn sympy pandas pillow

      2. -
    -

    For Python3, replace pip with pip3.

    -

    For OSX users we recommend, after having installed Xcode, to -install brew. Brew allows for a seamless installation of additional -software via for example

    -
      -

    1. brew install python3

      2. -
    -

    For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution, -you can use pip as well and simply install Python as

    -
      -

    1. sudo apt-get install python3 (or python for pyhton2.7)

      2. -
    -

    etc etc.

    -
    -
    -

    3.4. Python installers

    -

    If you don’t want to perform these operations separately and venture -into the hassle of exploring how to set up dependencies and paths, we -recommend two widely used distrubutions which set up all relevant -dependencies for Python, namely

    +whether we deal with supervised or unsupervised learning.

      -

    • Anaconda,

      • +

    • The first ingredient is normally our data set (which can be subdivided into training, validation and test data). Many find the most difficult part of using Machine Learning to be the set up of your data in a meaningful way.

      • +

    • The second item is a model which is normally a function of some parameters. The model reflects our knowledge of the system (or lack thereof). As an example, if we know that our data show a behavior similar to what would be predicted by a polynomial, fitting our data to a polynomial of some degree would then determin our model.

      • +

    • The last ingredient is a so-called cost/loss function (or error or risk function) which allows us to present an estimate on how good our model is in reproducing the data it is supposed to train.

    -

    which is an open source -distribution of the Python and R programming languages for large-scale -data processing, predictive analytics, and scientific computing, that -aims to simplify package management and deployment. Package versions -are managed by the package management system conda.

    - -

    is a Python -distribution for scientific and analytic computing distribution and -analysis environment, available for free and under a commercial -license.

    -

    Furthermore, Google’s Colab is a free Jupyter notebook environment that requires -no setup and runs entirely in the cloud. Try it out!

    +

    At the heart of basically all Machine Learning algorithms we will encounter so-called minimization or optimization algorithms. A large family of such methods are so-called gradient methods.

    +
    +

    3.2.1. A Frequentist approach to data analysis

    +

    When you hear phrases like predictions and estimations and +correlations and causations, what do you think of? May be you think +of the difference between classifying new data points and generating +new data points. +Or perhaps you consider that correlations represent some kind of symmetric statements like +if \(A\) is correlated with \(B\), then \(B\) is correlated with +\(A\). Causation on the other hand is directional, that is if \(A\) causes \(B\), \(B\) does not +necessarily cause \(A\).

    +

    These concepts are in some sense the difference between machine +learning and statistics. In machine learning and prediction based +tasks, we are often interested in developing algorithms that are +capable of learning patterns from given data in an automated fashion, +and then using these learned patterns to make predictions or +assessments of newly given data. In many cases, our primary concern +is the quality of the predictions or assessments, and we are less +concerned about the underlying patterns that were learned in order +to make these predictions.

    +

    In machine learning we normally use a so-called frequentist approach, +where the aim is to make predictions and find correlations. We focus +less on for example extracting a probability distribution function (PDF). The PDF can be +used in turn to make estimations and find causations such as given \(A\) +what is the likelihood of finding \(B\).

    +
    +
    +

    3.2.2. What is a good model?

    +

    In science and engineering we often end up in situations where we want to infer (or learn) a +quantitative model \(M\) for a given set of sample points \(\boldsymbol{X} \in [x_1, x_2,\dots x_N]\).

    +

    As we will see repeatedely in these lectures, we could try to fit these data points to a model given by a +straight line, or if we wish to be more sophisticated to a more complex +function.

    +

    The reason for inferring such a model is that it +serves many useful purposes. On the one hand, the model can reveal information +encoded in the data or underlying mechanisms from which the data were generated. For instance, we could discover important +corelations that relate interesting physics interpretations.

    +

    In addition, it can simplify the representation of the given data set and help +us in making predictions about future data samples.

    +

    A first important consideration to keep in mind is that inferring the correct model +for a given data set is an elusive, if not impossible, task. The fundamental difficulty +is that if we are not specific about what we mean by a correct model, there +could easily be many different models that fit the given data set equally well.

    +

    The central question is this: what leads us to say that a model is correct or +optimal for a given data set? To make the model inference problem well posed, i.e., +to guarantee that there is a unique optimal model for the given data, we need to +impose additional assumptions or restrictions on the class of models considered. To +this end, we should not be looking for just any model that can describe the data. +Instead, we should look for a model \(M\) that is the best among a restricted class +of models. In addition, to make the model inference problem computationally +tractable, we need to specify how restricted the class of models needs to be. A +common strategy is to start +with the simplest possible class of models that is just necessary to describe the data +or solve the problem at hand. More precisely, the model class should be rich enough +to contain at least one model that can fit the data to a desired accuracy and yet be +restricted enough that it is relatively simple to find the best model for the given data.

    +

    Thus, the most popular strategy is to start from the +simplest class of models and increase the complexity of the models only when the +simpler models become inadequate. For instance, if we work with a regression problem to fit a set of sample points, one +may first try the simplest class of models, namely linear models, followed obviously by more complex models.

    +

    How to evaluate which model fits best the data is something we will come back to over and over again in these sets of lectures.

    -
    -

    3.5. Useful Python libraries

    -

    Here we list several useful Python libraries we strongly recommend (if you use anaconda many of these are already there)

    -
      -

    • NumPy is a highly popular library for large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays

      • -

    • The pandas library provides high-performance, easy-to-use data structures and data analysis tools

      • -

    • Xarray is a Python package that makes working with labelled multi-dimensional arrays simple, efficient, and fun!

      • -

    • Scipy (pronounced “Sigh Pie”) is a Python-based ecosystem of open-source software for mathematics, science, and engineering.

      • -

    • Matplotlib is a Python 2D plotting library which produces publication quality figures in a variety of hardcopy formats and interactive environments across platforms.

      • -

    • Autograd can automatically differentiate native Python and Numpy code. It can handle a large subset of Python’s features, including loops, ifs, recursion and closures, and it can even take derivatives of derivatives of derivatives

      • -

    • SymPy is a Python library for symbolic mathematics.

      • -

    • scikit-learn has simple and efficient tools for machine learning, data mining and data analysis

      • -

    • TensorFlow is a Python library for fast numerical computing created and released by Google

      • -

    • Keras is a high-level neural networks API, written in Python and capable of running on top of TensorFlow, CNTK, or Theano

      • -

    • And many more such as pytorch, Theano etc

      • -
    -
    -
    -

    3.6. Installing R, C++, cython or Julia

    -

    You will also find it convenient to utilize R. We will mainly -use Python during our lectures and in various projects and exercises. -Those of you -already familiar with R should feel free to continue using R, keeping -however an eye on the parallel Python set ups. Similarly, if you are a -Python afecionado, feel free to explore R as well. Jupyter/Ipython -notebook allows you to run R codes interactively in your -browser. The software library R is really tailored for statistical data analysis -and allows for an easy usage of the tools and algorithms we will discuss in these -lectures.

    -

    To install R with Jupyter notebook -follow the link here

    -
    -
    -

    3.7. Installing R, C++, cython, Numba etc

    -

    For the C++ aficionados, Jupyter/IPython notebook allows you also to -install C++ and run codes written in this language interactively in -the browser. Since we will emphasize writing many of the algorithms -yourself, you can thus opt for either Python or C++ (or Fortran or other compiled languages) as programming -languages.

    -

    To add more entropy, cython can also be used when running your -notebooks. It means that Python with the jupyter notebook -setup allows you to integrate widely popular softwares and tools for -scientific computing. Similarly, the -Numba Python package delivers increased performance -capabilities with minimal rewrites of your codes. With its -versatility, including symbolic operations, Python offers a unique -computational environment. Your jupyter notebook can easily be -converted into a nicely rendered PDF file or a Latex file for -further processing. For example, convert to latex as

    -
        pycod jupyter nbconvert filename.ipynb --to latex 
-
    -
    -

    And to add more versatility, the Python package SymPy is a Python library for symbolic mathematics. It aims to become a full-featured computer algebra system (CAS) and is entirely written in Python.

    -

    Finally, if you wish to use the light mark-up language -doconce you can convert a standard ascii text file into various HTML -formats, ipython notebooks, latex files, pdf files etc with minimal edits. These lectures were generated using doconce.

    -
    -
    -

    3.8. Numpy examples and Important Matrix and vector handling packages

    -

    There are several central software libraries for linear algebra and eigenvalue problems. Several of the more -popular ones have been wrapped into ofter software packages like those from the widely used text Numerical Recipes. The original source codes in many of the available packages are often taken from the widely used -software package LAPACK, which follows two other popular packages -developed in the 1970s, namely EISPACK and LINPACK. We describe them shortly here.

    -
      -

    • LINPACK: package for linear equations and least square problems.

      • -

    • LAPACK:package for solving symmetric, unsymmetric and generalized eigenvalue problems. From LAPACK’s website http://www.netlib.org it is possible to download for free all source codes from this library. Both C/C++ and Fortran versions are available.

      • -

    • BLAS (I, II and III): (Basic Linear Algebra Subprograms) are routines that provide standard building blocks for performing basic vector and matrix operations. Blas I is vector operations, II vector-matrix operations and III matrix-matrix operations. Highly parallelized and efficient codes, all available for download from http://www.netlib.org.

      • -
    -
    -
    -

    3.9. Basic Matrix Features

    -

    Matrix properties reminder

    -
    -\[\begin{split} -\mathbf{A} = - \begin{bmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\ - a_{21} & a_{22} & a_{23} & a_{24} \\ - a_{31} & a_{32} & a_{33} & a_{34} \\ - a_{41} & a_{42} & a_{43} & a_{44} - \end{bmatrix}\qquad -\mathbf{I} = - \begin{bmatrix} 1 & 0 & 0 & 0 \\ - 0 & 1 & 0 & 0 \\ - 0 & 0 & 1 & 0 \\ - 0 & 0 & 0 & 1 - \end{bmatrix} -\end{split}\]
    -

    The inverse of a matrix is defined by

    -
    -\[ -\mathbf{A}^{-1} \cdot \mathbf{A} = I -\]
    - - - - - - - - - - - -
    Relations Name matrix elements
    $A = A^{T}$ symmetric $a_{ij} = a_{ji}$
    $A = \left (A^{T} \right )^{-1}$ real orthogonal $\sum_k a_{ik} a_{jk} = \sum_k a_{ki} a_{kj} = \delta_{ij}$
    $A = A^{ * }$ real matrix $a_{ij} = a_{ij}^{ * }$
    $A = A^{\dagger}$ hermitian $a_{ij} = a_{ji}^{ * }$
    $A = \left (A^{\dagger} \right )^{-1}$ unitary $\sum_k a_{ik} a_{jk}^{ * } = \sum_k a_{ki}^{ * } a_{kj} = \delta_{ij}$
    -
    -

    3.9.1. Some famous Matrices

    -
      -

    • Diagonal if \(a_{ij}=0\) for \(i\ne j\)

      • -

    • Upper triangular if \(a_{ij}=0\) for \(i > j\)

      • -

    • Lower triangular if \(a_{ij}=0\) for \(i < j\)

      • -

    • Upper Hessenberg if \(a_{ij}=0\) for \(i > j+1\)

      • -

    • Lower Hessenberg if \(a_{ij}=0\) for \(i < j+1\)

      • -

    • Tridiagonal if \(a_{ij}=0\) for \(|i -j| > 1\)

      • -

    • Lower banded with bandwidth \(p\): \(a_{ij}=0\) for \(i > j+p\)

      • -

    • Upper banded with bandwidth \(p\): \(a_{ij}=0\) for \(i < j+p\)

      • -

    • Banded, block upper triangular, block lower triangular….

      • -
    -
    -
    -

    3.9.2. More Basic Matrix Features

    -

    Some Equivalent Statements -For an \(N\times N\) matrix \(\mathbf{A}\) the following properties are all equivalent

    -
      -

    • If the inverse of \(\mathbf{A}\) exists, \(\mathbf{A}\) is nonsingular.

      • -

    • The equation \(\mathbf{Ax}=0\) implies \(\mathbf{x}=0\).

      • -

    • The rows of \(\mathbf{A}\) form a basis of \(R^N\).

      • -

    • The columns of \(\mathbf{A}\) form a basis of \(R^N\).

      • -

    • \(\mathbf{A}\) is a product of elementary matrices.

      • -

    • \(0\) is not eigenvalue of \(\mathbf{A}\).

      • -
    -
    -
    -
    -

    3.10. Numpy and arrays

    -

    Numpy provides an easy way to handle arrays in Python. The standard way to import this library is as

    -
    -
    -
    import numpy as np
-
    -
    -
    -
    -

    Here follows a simple example where we set up an array of ten elements, all determined by random numbers drawn according to the normal distribution,

    -
    -
    -
    n = 10
-x = np.random.normal(size=n)
-print(x)
-
    -
    -
    -
    -
    [ 0.03607832  0.57781668 -2.61124978 -0.08271336 -1.17654307  2.45019484
-  1.40902095 -0.07159175 -0.52722156 -0.83614019]
-
    -
    -
    -
    -

    We defined a vector \(x\) with \(n=10\) elements with its values given by the Normal distribution \(N(0,1)\). -Another alternative is to declare a vector as follows

    -
    -
    -
    import numpy as np
-x = np.array([1, 2, 3])
-print(x)
-
    -
    -
    -
    -
    [1 2 3]
-
    -
    -
    -
    -

    Here we have defined a vector with three elements, with \(x_0=1\), \(x_1=2\) and \(x_2=3\). Note that both Python and C++ -start numbering array elements from \(0\) and on. This means that a vector with \(n\) elements has a sequence of entities \(x_0, x_1, x_2, \dots, x_{n-1}\). We could also let (recommended) Numpy to compute the logarithms of a specific array as

    -
    -
    -
    import numpy as np
-x = np.log(np.array([4, 7, 8]))
-print(x)
-
    -
    -
    -
    -
    [1.38629436 1.94591015 2.07944154]
-
    -
    -
    -
    -

    In the last example we used Numpy’s unary function \(np.log\). This function is -highly tuned to compute array elements since the code is vectorized -and does not require looping. We normaly recommend that you use the -Numpy intrinsic functions instead of the corresponding log function -from Python’s math module. The looping is done explicitely by the -np.log function. The alternative, and slower way to compute the -logarithms of a vector would be to write

    -
    -
    -
    import numpy as np
-from math import log
-x = np.array([4, 7, 8])
-for i in range(0, len(x)):
-    x[i] = log(x[i])
-print(x)
-
    -
    -
    -
    -
    [1 1 2]
-
    -
    -
    -
    -

    We note that our code is much longer already and we need to import the log function from the math module. -The attentive reader will also notice that the output is \([1, 1, 2]\). Python interprets automagically our numbers as integers (like the automatic keyword in C++). To change this we could define our array elements to be double precision numbers as

    -
    -
    -
    import numpy as np
-x = np.log(np.array([4, 7, 8], dtype = np.float64))
-print(x)
-
    -
    -
    -
    -
    [1.38629436 1.94591015 2.07944154]
-
    -
    -
    -
    -

    or simply write them as double precision numbers (Python uses 64 bits as default for floating point type variables), that is

    -
    -
    -
    import numpy as np
-x = np.log(np.array([4.0, 7.0, 8.0])
-print(x)
-
    -
    -
    -
    -
      File "<ipython-input-7-f6d7a289d493>", line 3
-    print(x)
-    ^
-SyntaxError: invalid syntax
-
    -
    -
    -
    -

    To check the number of bytes (remember that one byte contains eight bits for double precision variables), you can use simple use the itemsize functionality (the array \(x\) is actually an object which inherits the functionalities defined in Numpy) as

    -
    -
    -
    import numpy as np
-x = np.log(np.array([4.0, 7.0, 8.0])
-print(x.itemsize)
-
    -
    -
    -
    -
    -
    -

    3.11. Matrices in Python

    -

    Having defined vectors, we are now ready to try out matrices. We can -define a \(3 \times 3 \) real matrix \(\hat{A}\) as (recall that we user -lowercase letters for vectors and uppercase letters for matrices)

    -
    -
    -
    import numpy as np
-A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))
-print(A)
-
    -
    -
    -
    -

    If we use the shape function we would get \((3, 3)\) as output, that is verifying that our matrix is a \(3\times 3\) matrix. We can slice the matrix and print for example the first column (Python organized matrix elements in a row-major order, see below) as

    -
    -
    -
    import numpy as np
-A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))
-# print the first column, row-major order and elements start with 0
-print(A[:,0])
-
    -
    -
    -
    -

    We can continue this was by printing out other columns or rows. The example here prints out the second column

    -
    -
    -
    import numpy as np
-A = np.log(np.array([ [4.0, 7.0, 8.0], [3.0, 10.0, 11.0], [4.0, 5.0, 7.0] ]))
-# print the first column, row-major order and elements start with 0
-print(A[1,:])
-
    -
    -
    -
    -

    Numpy contains many other functionalities that allow us to slice, subdivide etc etc arrays. We strongly recommend that you look up the Numpy website for more details. Useful functions when defining a matrix are the np.zeros function which declares a matrix of a given dimension and sets all elements to zero

    -
    -
    -
    import numpy as np
-n = 10
-# define a matrix of dimension 10 x 10 and set all elements to zero
-A = np.zeros( (n, n) )
-print(A)
-
    -
    -
    -
    -

    or initializing all elements to

    -
    -
    -
    import numpy as np
-n = 10
-# define a matrix of dimension 10 x 10 and set all elements to one
-A = np.ones( (n, n) )
-print(A)
-
    -
    -
    -
    -

    or as unitarily distributed random numbers (see the material on random number generators in the statistics part)

    -
    -
    -
    import numpy as np
-n = 10
-# define a matrix of dimension 10 x 10 and set all elements to random numbers with x \in [0, 1]
-A = np.random.rand(n, n)
-print(A)
-
    -
    -
    -
    -

    As we will see throughout these lectures, there are several extremely useful functionalities in Numpy. -As an example, consider the discussion of the covariance matrix. Suppose we have defined three vectors -\(\hat{x}, \hat{y}, \hat{z}\) with \(n\) elements each. The covariance matrix is defined as

    -
    -\[\begin{split} -\hat{\Sigma} = \begin{bmatrix} \sigma_{xx} & \sigma_{xy} & \sigma_{xz} \\ - \sigma_{yx} & \sigma_{yy} & \sigma_{yz} \\ - \sigma_{zx} & \sigma_{zy} & \sigma_{zz} - \end{bmatrix}, -\end{split}\]
    -

    where for example

    -
    -\[ -\sigma_{xy} =\frac{1}{n} \sum_{i=0}^{n-1}(x_i- \overline{x})(y_i- \overline{y}). -\]
    -

    The Numpy function np.cov calculates the covariance elements using the factor \(1/(n-1)\) instead of \(1/n\) since it assumes we do not have the exact mean values. -The following simple function uses the np.vstack function which takes each vector of dimension \(1\times n\) and produces a \(3\times n\) matrix \(\hat{W}\)

    -
    -\[\begin{split} -\hat{W} = \begin{bmatrix} x_0 & y_0 & z_0 \\ - x_1 & y_1 & z_1 \\ - x_2 & y_2 & z_2 \\ - \dots & \dots & \dots \\ - x_{n-2} & y_{n-2} & z_{n-2} \\ - x_{n-1} & y_{n-1} & z_{n-1} - \end{bmatrix}, -\end{split}\]
    -

    which in turn is converted into into the \(3\times 3\) covariance matrix -\(\hat{\Sigma}\) via the Numpy function np.cov(). We note that we can also calculate -the mean value of each set of samples \(\hat{x}\) etc using the Numpy -function np.mean(x). We can also extract the eigenvalues of the -covariance matrix through the np.linalg.eig() function.

    -
    -
    -
    # Importing various packages
-import numpy as np
-
-n = 100
-x = np.random.normal(size=n)
-print(np.mean(x))
-y = 4+3*x+np.random.normal(size=n)
-print(np.mean(y))
-z = x**3+np.random.normal(size=n)
-print(np.mean(z))
-W = np.vstack((x, y, z))
-Sigma = np.cov(W)
-print(Sigma)
-Eigvals, Eigvecs = np.linalg.eig(Sigma)
-print(Eigvals)
-
    -
    -
    -
    -
    -
    -
    %matplotlib inline
-
-import numpy as np
-import matplotlib.pyplot as plt
-from scipy import sparse
-eye = np.eye(4)
-print(eye)
-sparse_mtx = sparse.csr_matrix(eye)
-print(sparse_mtx)
-x = np.linspace(-10,10,100)
-y = np.sin(x)
-plt.plot(x,y,marker='x')
-plt.show()
-
    -
    -
    -
    -
    -
    -

    3.12. Meet the Pandas

    - -

    Another useful Python package is -pandas, which is an open source library -providing high-performance, easy-to-use data structures and data -analysis tools for Python. pandas stands for panel data, a term borrowed from econometrics and is an efficient library for data analysis with an emphasis on tabular data. -pandas has two major classes, the DataFrame class with two-dimensional data objects and tabular data organized in columns and the class Series with a focus on one-dimensional data objects. Both classes allow you to index data easily as we will see in the examples below. -pandas allows you also to perform mathematical operations on the data, spanning from simple reshapings of vectors and matrices to statistical operations.

    -

    The following simple example shows how we can, in an easy way make tables of our data. Here we define a data set which includes names, place of birth and date of birth, and displays the data in an easy to read way. We will see repeated use of pandas, in particular in connection with classification of data.

    -
    -
    -
    import pandas as pd
-from IPython.display import display
-data = {'First Name': ["Frodo", "Bilbo", "Aragorn II", "Samwise"],
-        'Last Name': ["Baggins", "Baggins","Elessar","Gamgee"],
-        'Place of birth': ["Shire", "Shire", "Eriador", "Shire"],
-        'Date of Birth T.A.': [2968, 2890, 2931, 2980]
-        }
-data_pandas = pd.DataFrame(data)
-display(data_pandas)
-
    -
    -
    -
    -

    In the above we have imported pandas with the shorthand pd, the latter has become the standard way we import pandas. We make then a list of various variables -and reorganize the aboves lists into a DataFrame and then print out a neat table with specific column labels as Name, place of birth and date of birth. -Displaying these results, we see that the indices are given by the default numbers from zero to three. -pandas is extremely flexible and we can easily change the above indices by defining a new type of indexing as

    -
    -
    -
    data_pandas = pd.DataFrame(data,index=['Frodo','Bilbo','Aragorn','Sam'])
-display(data_pandas)
-
    -
    -
    -
    -

    Thereafter we display the content of the row which begins with the index Aragorn

    -
    -
    -
    display(data_pandas.loc['Aragorn'])
-
    -
    -
    -
    -

    We can easily append data to this, for example

    -
    -
    -
    new_hobbit = {'First Name': ["Peregrin"],
-              'Last Name': ["Took"],
-              'Place of birth': ["Shire"],
-              'Date of Birth T.A.': [2990]
-              }
-data_pandas=data_pandas.append(pd.DataFrame(new_hobbit, index=['Pippin']))
-display(data_pandas)
-
    -
    -
    -
    -

    Here are other examples where we use the DataFrame functionality to handle arrays, now with more interesting features for us, namely numbers. We set up a matrix -of dimensionality \(10\times 5\) and compute the mean value and standard deviation of each column. Similarly, we can perform mathematial operations like squaring the matrix elements and many other operations.

    -
    -
    -
    import numpy as np
-import pandas as pd
-from IPython.display import display
-np.random.seed(100)
-# setting up a 10 x 5 matrix
-rows = 10
-cols = 5
-a = np.random.randn(rows,cols)
-df = pd.DataFrame(a)
-display(df)
-print(df.mean())
-print(df.std())
-display(df**2)
-
    -
    -
    -
    -

    Thereafter we can select specific columns only and plot final results

    -
    -
    -
    df.columns = ['First', 'Second', 'Third', 'Fourth', 'Fifth']
-df.index = np.arange(10)
-
-display(df)
-print(df['Second'].mean() )
-print(df.info())
-print(df.describe())
-
-from pylab import plt, mpl
-plt.style.use('seaborn')
-mpl.rcParams['font.family'] = 'serif'
-
-df.cumsum().plot(lw=2.0, figsize=(10,6))
-plt.show()
-
-
-df.plot.bar(figsize=(10,6), rot=15)
-plt.show()
-
    -
    -
    -
    -

    We can produce a \(4\times 4\) matrix

    -
    -
    -
    b = np.arange(16).reshape((4,4))
-print(b)
-df1 = pd.DataFrame(b)
-print(df1)
-
    -
    -
    -
    -

    and many other operations.

    -

    The Series class is another important class included in -pandas. You can view it as a specialization of DataFrame but where -we have just a single column of data. It shares many of the same features as _DataFrame. As with DataFrame, -most operations are vectorized, achieving thereby a high performance when dealing with computations of arrays, in particular labeled arrays. -As we will see below it leads also to a very concice code close to the mathematical operations we may be interested in. -For multidimensional arrays, we recommend strongly xarray. xarray has much of the same flexibility as pandas, but allows for the extension to higher dimensions than two. We will see examples later of the usage of both pandas and xarray.

    -

    In order to study various Machine Learning algorithms, we need to -access data. Acccessing data is an essential step in all machine -learning algorithms. In particular, setting up the so-called design -matrix (to be defined below) is often the first element we need in -order to perform our calculations. To set up the design matrix means -reading (and later, when the calculations are done, writing) data -in various formats, The formats span from reading files from disk, -loading data from databases and interacting with online sources -like web application programming interfaces (APIs).

    -

    In handling various input formats, as discussed above, we will mainly stay with pandas, -a Python package which allows us, in a seamless and painless way, to -deal with a multitude of formats, from standard csv (comma separated -values) files, via excel, html to hdf5 formats. With pandas -and the DataFrame and Series functionalities we are able to convert text data -into the calculational formats we need for a specific algorithm. And our code is going to be -pretty close the basic mathematical expressions.

    -

    Our first data set is going to be a classic from nuclear physics, namely all -available data on binding energies. Don’t be intimidated if you are not familiar with nuclear physics. It serves simply as an example here of a data set.

    -

    We will show some of the -strengths of packages like Scikit-Learn in fitting nuclear binding energies to -specific functions using linear regression first. Then, as a teaser, we will show you how -you can easily implement other algorithms like decision trees and random forests and neural networks.

    -

    But before we really start with nuclear physics data, let’s just look at some simpler polynomial fitting cases, such as, -(don’t be offended) fitting straight lines!

    -

    3.13. Simple linear regression model using scikit-learn

    +

    3.3. Simple linear regression model using scikit-learn

    We start with perhaps our simplest possible example, using Scikit-Learn to perform linear regression analysis on a data set produced by us.

    What follows is a simple Python code where we have defined a function \(y\) in terms of the variable \(x\). Both are defined as vectors with \(100\) entries. @@ -1198,7 +625,9 @@ data with a straight line.

    The Python code follows here.

    -
    # Importing various packages
+
    %matplotlib inline
+
+# Importing various packages
 import numpy as np
 import matplotlib.pyplot as plt
 from sklearn.linear_model import LinearRegression
@@ -1220,6 +649,9 @@ data with a straight line.
    +
    +_images/chapter1_3_0.png +

    This example serves several aims. It allows us to demonstrate several aspects of data analysis and later machine learning algorithms. The @@ -1301,6 +733,9 @@ to be dominated by outliers.

    +
    +_images/chapter1_11_0.png +

    Depending on the parameter in front of the normal distribution, we may have a small or larger relative error. Try to play around with @@ -1344,6 +779,19 @@ example of the functionality of Scikit-Learn.

    +
    +
    The intercept alpha: 
+ [2.18954529]
+Coefficient beta : 
+ [[4.57051369]]
+Mean squared error: 0.23
+Variance score: 0.88
+Mean squared log error: 0.01
+Mean absolute error: 0.39
+
    +
    +_images/chapter1_13_1.png +

    The function coef gives us the parameter \(\beta\) of our fit while intercept yields \(\alpha\). Depending on the constant in front of the normal distribution, we get values near or far from \(alpha =2\) and \(\beta =5\). Try to play around with different parameters in front of the normal distribution. The function meansquarederror gives us the mean square error, a risk metric corresponding to the expected value of the squared (quadratic) error or loss defined as

    @@ -1396,7 +844,8 @@ ways of dealing with outliers.

    The Huber cost function is defined as

    \[\begin{split} -H_{\delta}(a)={\begin{cases}{\frac {1}{2}}{a^{2}}&{\text{for }}|a|\leq \delta ,\\\delta (|a|-{\frac {1}{2}}\delta ),&{\text{otherwise.}}\end{cases}}}. +H_{\delta}(a)=\left\{\begin{array}\frac{1}{2}a^{2}&{\text{for }}|a|\leq \delta ,\\ \delta (|a|-\frac {1}{2}\del\ +ta ),&{\text{otherwise.}\end{array}\right. \end{split}\]

    Here \(a=\boldsymbol{y} - \boldsymbol{\tilde{y}}\). We will discuss in more @@ -1437,6 +886,12 @@ a linear \(x\)-dependence we s +

    +_images/chapter1_27_0.png +
    0.004999999999999997
+
    +
    +

    Let us now dive into nuclear physics and remind ourselves briefly about some basic features about binding energies. A basic quantity which can be measured for the ground @@ -1499,7 +954,7 @@ to the experimental data.

    arises from the tendency of proton pairs and neutron pairs to occur. An even number of particles is more stable than an odd number.

    -

    3.13.1. Organizing our data

    +

    3.3.1. Organizing our data

    Let us start with reading and organizing our data. We start with the compilation of masses and binding energies from 2016. After having downloaded this file to our own computer, we are now ready to read the file and start structuring our data.

    @@ -1542,6 +997,18 @@ After having downloaded this file to our own computer, we are now ready to read
    +
    +
    ---------------------------------------------------------------------------
+FileNotFoundError                         Traceback (most recent call last)
+<ipython-input-5-3cd19a0768e1> in <module>
+     31     plt.savefig(image_path(fig_id) + ".png", format='png')
+     32 
+---> 33 infile = open(data_path("MassEval2016.dat"),'r')
+
+FileNotFoundError: [Errno 2] No such file or directory: 'DataFiles/MassEval2016.dat'
+
    +
    +

    Before we proceed, we define also a function for making our plots. You can obviously avoid this and simply set up various matplotlib commands every time you need them. You may however find it convenient to collect all such commands in one function and simply call this function.

    @@ -1759,8 +1226,8 @@ functionality.

    -
    -

    3.14. Linear Regression, basic elements

    +
    +

    3.4. Linear Regression, basic elements

    Video of Lecture.

    Fitting a continuous function with linear parameterization in terms of the parameters \(\boldsymbol{\beta}\).

      @@ -2091,7 +1558,7 @@ allow for the usage of direct linear algebra methods such as LU

      The following matrix and vector relation will be useful here and for the rest of the course. Vectors are always written as boldfaced lower case letters and matrices as upper case boldfaced letters.

      4 -8

      +3

      < < < @@ -2108,7 +1575,7 @@ O C K

      4 -9

      +4

      < < < @@ -2124,8 +1591,8 @@ L O C K

      -

      5 -0

      +

      4 +5

      < < < @@ -2243,7 +1710,7 @@ Since we are not using Scikit-Learn here we can define our own

    -

    3.14.1. The \(\chi^2\) function

    +

    3.4.1. The \(\chi^2\) function

    Normally, the response (dependent or outcome) variable \(y_i\) is the outcome of a numerical experiment or another type of experiment and is thus only an approximation to the true value. It is then always @@ -2363,7 +1830,7 @@ Singular Value Decomposition (SVD) method discussed below. Or using Lasso and Ridge regression. See below.

    -

    3.14.2. Fitting an Equation of State for Dense Nuclear Matter

    +

    3.4.2. Fitting an Equation of State for Dense Nuclear Matter

    Before we continue, let us introduce yet another example. We are going to fit the nuclear equation of state using results from many-body calculations. The equation of state we have made available here, as function of @@ -2477,7 +1944,7 @@ below.

    -

    3.15. Splitting our Data in Training and Test data

    +

    3.5. Splitting our Data in Training and Test data

    It is normal in essentially all Machine Learning studies to split the data in a training set and a test set (sometimes also an additional validation set). Scikit-Learn has an own function for this. There @@ -2560,7 +2027,7 @@ but now splitting the data into a training set and a test set.

    -

    3.16. The Boston housing data example

    +

    3.6. The Boston housing data example

    The Boston housing
    data set was originally a part of UCI Machine Learning Repository and has been removed now. The data set is now included in Scikit-Learn’s @@ -2585,7 +2052,7 @@ the house using the features (predictors) listed here.

    -

    3.17. Housing data, the code

    +

    3.7. Housing data, the code

    We start by importing the libraries

    @@ -2751,7 +2218,7 @@ the house using the features (predictors) listed here.

    -

    3.18. Reducing the number of degrees of freedom, overarching view

    +

    3.8. Reducing the number of degrees of freedom, overarching view

    Many Machine Learning problems involve thousands or even millions of features for each training instance. Not only does this make training extremely slow, it can also make it much harder to find a good @@ -2799,7 +2266,7 @@ RobustScaler ignore data points that are very different from the rest outliers, and might often lead to trouble for other scaling techniques.

    -

    3.18.1. Simple preprocessing examples, Franke function and regression

    +

    3.8.1. Simple preprocessing examples, Franke function and regression

    # Common imports
@@ -2900,6 +2367,233 @@ techniques.
    +
    +

    3.9. Exercises

    +
    +

    3.9.1. Exercise: Setting up various Python environments

    +

    The first exercise here is of a mere technical art. We want you to have

    +
      +

    • git as a version control software and to establish a user account on a provider like GitHub. Other providers like GitLab etc are equally fine. You can also use the University of Oslo GitHub facilities.

      • +

    • Install various Python packages

      • +
    +

    We will make extensive use of Python as programming language and its +myriad of available libraries. You will find +IPython/Jupyter notebooks invaluable in your work. You can run R +codes in the Jupyter/IPython notebooks, with the immediate benefit of +visualizing your data. You can also use compiled languages like C++, +Rust, Fortran etc if you prefer. The focus in these lectures will be +on Python.

    +

    If you have Python installed (we recommend Python3) and you feel +pretty familiar with installing different packages, we recommend that +you install the following Python packages via pip as

    +
      +

    1. pip install numpy scipy matplotlib ipython scikit-learn sympy pandas pillow

      2. +
    +

    For Tensorflow, we recommend following the instructions in the text of +Aurelien Geron, Hands‑On Machine Learning with Scikit‑Learn and TensorFlow, O’Reilly

    +

    We will come back to tensorflow later.

    +

    For Python3, replace pip with pip3.

    +

    For OSX users we recommend, after having installed Xcode, to +install brew. Brew allows for a seamless installation of additional +software via for example

    +
      +

    1. brew install python3

      2. +
    +

    For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution, +you can use pip as well and simply install Python as

    +
      +

    1. sudo apt-get install python3 (or python for Python2.7)

      2. +
    +

    If you don’t want to perform these operations separately and venture +into the hassle of exploring how to set up dependencies and paths, we +recommend two widely used distrubutions which set up all relevant +dependencies for Python, namely

    + +

    which is an open source +distribution of the Python and R programming languages for large-scale +data processing, predictive analytics, and scientific computing, that +aims to simplify package management and deployment. Package versions +are managed by the package management system conda.

    + +

    is a Python +distribution for scientific and analytic computing distribution and +analysis environment, available for free and under a commercial +license.

    +

    We recommend using Anaconda if you are not too familiar with setting paths in a terminal environment.

    +
    +
    +

    3.9.2. Exercise: making your own data and exploring scikit-learn

    +

    We will generate our own dataset for a function \(y(x)\) where \(x \in [0,1]\) and defined by random numbers computed with the uniform distribution. The function \(y\) is a quadratic polynomial in \(x\) with added stochastic noise according to the normal distribution \(\cal {N}(0,1)\). +The following simple Python instructions define our \(x\) and \(y\) values (with 100 data points).

    +
    +
    +
    x = np.random.rand(100,1)
+y = 2.0+5*x*x+0.1*np.random.randn(100,1)
+
    +
    +
    +
    +
      +

    1. Write your own code (following the examples under the regression notes) for computing the parametrization of the data set fitting a second-order polynomial.

      2. +

    2. Use thereafter scikit-learn (see again the examples in the regression slides) and compare with your own code.

      3. +

    3. Using scikit-learn, compute also the mean square error, a risk metric corresponding to the expected value of the squared (quadratic) error defined as

      4. +
    +
    +\[ +MSE(\hat{y},\hat{\tilde{y}}) = \frac{1}{n} +\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2, +\]
    +

    and the \(R^2\) score function. +If \(\tilde{\hat{y}}_i\) is the predicted value of the \(i-th\) sample and \(y_i\) is the corresponding true value, then the score \(R^2\) is defined as

    +
    +\[ +R^2(\hat{y}, \tilde{\hat{y}}) = 1 - \frac{\sum_{i=0}^{n - 1} (y_i - \tilde{y}_i)^2}{\sum_{i=0}^{n - 1} (y_i - \bar{y})^2}, +\]
    +

    where we have defined the mean value of \(\hat{y}\) as

    +
    +\[ +\bar{y} = \frac{1}{n} \sum_{i=0}^{n - 1} y_i. +\]
    +

    You can use the functionality included in scikit-learn. If you feel for it, you can use your own program and define functions which compute the above two functions. +Discuss the meaning of these results. Try also to vary the coefficient in front of the added stochastic noise term and discuss the quality of the fits.

    +

    !bsol +The code here is an example of where we define our own design matrix and fit parameters \(\beta\).

    +
    +
    +
    import os
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+from sklearn.model_selection import train_test_split
+
+def save_fig(fig_id):
+    plt.savefig(image_path(fig_id) + ".png", format='png')
+
+def R2(y_data, y_model):
+    return 1 - np.sum((y_data - y_model) ** 2) / np.sum((y_data - np.mean(y_data)) ** 2)
+def MSE(y_data,y_model):
+    n = np.size(y_model)
+    return np.sum((y_data-y_model)**2)/n
+
+x = np.random.rand(100)
+y = 2.0+5*x*x+0.1*np.random.randn(100)
+
+
+#  The design matrix now as function of a given polynomial
+X = np.zeros((len(x),3))
+X[:,0] = 1.0
+X[:,1] = x
+X[:,2] = x**2
+# We split the data in test and training data
+X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
+# matrix inversion to find beta
+beta = np.linalg.inv(X_train.T @ X_train) @ X_train.T @ y_train
+print(beta)
+# and then make the prediction
+ytilde = X_train @ beta
+print("Training R2")
+print(R2(y_train,ytilde))
+print("Training MSE")
+print(MSE(y_train,ytilde))
+ypredict = X_test @ beta
+print("Test R2")
+print(R2(y_test,ypredict))
+print("Test MSE")
+print(MSE(y_test,ypredict))
+
    +
    +
    +
    +

    !esol

    +
    +
    +

    3.9.3. Exercise: Normalizing our data

    +

    A much used approach before starting to train the data is to preprocess our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ.

    +

    Scikit-Learn has several functions which allow us to rescale the +data, normally resulting in much better results in terms of various +accuracy scores. The StandardScaler function in Scikit-Learn +ensures that for each feature/predictor we study the mean value is +zero and the variance is one (every column in the design/feature +matrix). This scaling has the drawback that it does not ensure that +we have a particular maximum or minimum in our data set. Another +function included in Scikit-Learn is the MinMaxScaler which +ensures that all features are exactly between \(0\) and \(1\). The

    +

    The Normalizer scales each data +point such that the feature vector has a euclidean length of one. In other words, it +projects a data point on the circle (or sphere in the case of higher dimensions) with a +radius of 1. This means every data point is scaled by a different number (by the +inverse of it’s length). +This normalization is often used when only the direction (or angle) of the data matters, +not the length of the feature vector.

    +

    The RobustScaler works similarly to the StandardScaler in that it +ensures statistical properties for each feature that guarantee that +they are on the same scale. However, the RobustScaler uses the median +and quartiles, instead of mean and variance. This makes the +RobustScaler ignore data points that are very different from the rest +(like measurement errors). These odd data points are also called +outliers, and might often lead to trouble for other scaling +techniques.

    +

    It also common to split the data in a training set and a testing set. A typical split is to use \(80\%\) of the data for training and the rest +for testing. This can be done as follows with our design matrix \(\boldsymbol{X}\) and data \(\boldsymbol{y}\) (remember to import scikit-learn)

    +
    +
    +
    # split in training and test data
+X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2)
+
    +
    +
    +
    +

    Then we can use the standard scaler to scale our data as

    +
    +
    +
    scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
    +
    +
    +
    +

    In this exercise we want you to to compute the MSE for the training +data and the test data as function of the complexity of a polynomial, +that is the degree of a given polynomial. We want you also to compute the \(R2\) score as function of the complexity of the model for both training data and test data. You should also run the calculation with and without scaling.

    +

    One of +the aims is to reproduce Figure 2.11 of Hastie et al.

    +

    Our data is defined by \(x\in [-3,3]\) with a total of for example \(100\) data points.

    +
    +
    +
    np.random.seed()
+n = 100
+maxdegree = 14
+# Make data set.
+x = np.linspace(-3, 3, n).reshape(-1, 1)
+y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)
+
    +
    +
    +
    +

    where \(y\) is the function we want to fit with a given polynomial. +!bsubex +Write a first code which sets up a design matrix \(X\) defined by a fifth-order polynomial. Scale your data and split it in training and test data. +!esubex

    +

    !bsubex +Perform an ordinary least squares and compute the means squared error and the \(R2\) factor for the training data and the test data, with and without scaling. +!esubex

    +

    !bsubex +Add now a model which allows you to make polynomials up to degree \(15\). Perform a standard OLS fitting of the training data and compute the MSE and \(R2\) for the training and test data and plot both test and training data MSE and \(R2\) as functions of the polynomial degree. Compare what you see with Figure 2.11 of Hastie et al. Comment your results. For which polynomial degree do you find an optimal MSE (smallest value)? +!bsol +Here you simply need to change the degree of the polynomial in the above code to \(n=15\). +!esol +!esubex

    +
    +