From f44378c5e4266788009853d1306c4d994d57b5fd Mon Sep 17 00:00:00 2001 From: mhjensen Date: Sat, 23 Nov 2019 23:00:49 +0100 Subject: [PATCH] correcting typos in svm --- doc/pub/svm/html/._svm-bs000.html | 2 +- doc/pub/svm/html/._svm-bs001.html | 6 ++++-- doc/pub/svm/html/._svm-bs003.html | 8 ++++++-- doc/pub/svm/html/svm-bs.html | 2 +- doc/pub/svm/html/svm-reveal.html | 18 ++++++++++++------ doc/pub/svm/html/svm-solarized.html | 16 +++++++++++----- doc/pub/svm/html/svm.html | 16 +++++++++++----- doc/pub/svm/ipynb/ipynb-svm-src.tar.gz | Bin 189 -> 189 bytes doc/pub/svm/ipynb/svm.ipynb | 14 +++++++++----- doc/pub/svm/pdf/svm-minted.pdf | Bin 266200 -> 266295 bytes doc/src/SupportVMachines/svm.do.txt | 14 ++++++++++---- 11 files changed, 65 insertions(+), 31 deletions(-) diff --git a/doc/pub/svm/html/._svm-bs000.html b/doc/pub/svm/html/._svm-bs000.html index 0ad4b19d0..b7903aa11 100644 --- a/doc/pub/svm/html/._svm-bs000.html +++ b/doc/pub/svm/html/._svm-bs000.html @@ -174,7 +174,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Nov 22, 2019

+

Nov 23, 2019


diff --git a/doc/pub/svm/html/._svm-bs001.html b/doc/pub/svm/html/._svm-bs001.html index bfb7ac6b8..6382d89df 100644 --- a/doc/pub/svm/html/._svm-bs001.html +++ b/doc/pub/svm/html/._svm-bs001.html @@ -159,7 +159,7 @@ MathJax.Hub.Config({

A Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning model, capable of performing linear or nonlinear +Machine Learning method, capable of performing linear or nonlinear classification, regression, and even outlier detection. It is one of the most popular models in Machine Learning, and anyone interested in Machine Learning should have it in their toolbox. SVMs are @@ -167,7 +167,9 @@ particularly well suited for classification of complex but small-sized or medium-sized datasets.

-The case with two well-separated classes only can be understood in an intuitive way in terms of lines in a two-dimensional space separating the two classes (see figure below). +The case with two well-separated classes only can be understood in an +intuitive way in terms of lines in a two-dimensional space separating +the two classes (see figure below).

The basic mathematics behind the SVM is however less familiar to most of us. diff --git a/doc/pub/svm/html/._svm-bs003.html b/doc/pub/svm/html/._svm-bs003.html index 4283e0279..f48b78c5a 100644 --- a/doc/pub/svm/html/._svm-bs003.html +++ b/doc/pub/svm/html/._svm-bs003.html @@ -158,7 +158,9 @@ MathJax.Hub.Config({

What is a hyperplane?

-The aim of the SVM algorithm is to find a hyperplane in an \( p \)-dimensional space, where \( p \) is the number of features that distinctly classifies the data points. +The aim of the SVM algorithm is to find a hyperplane in a +\( p \)-dimensional space, where \( p \) is the number of features that +distinctly classifies the data points.

In a \( p \)-dimensional space, a hyperplane is what we call an affine subspace of dimension of \( p-1 \). @@ -171,12 +173,14 @@ $$ b+w_1x_1+w_2x_2=0, $$ +

where \( b \) is the intercept and \( w_1 \) and \( w_2 \) define the elements of a vector orthogonal to the line \( b+w_1x_1+w_2x_2=0 \). In two dimensions we define the vectors \( \boldsymbol{x} =[x1,x2] \) and \( \boldsymbol{w}=[w1,w2] \). We can then rewrite the above equation as + $$ -\boldsymbol{w}^T\boldsymbol{x}+b=0. +\boldsymbol{x}^T\boldsymbol{w}+b=0. $$

diff --git a/doc/pub/svm/html/svm-bs.html b/doc/pub/svm/html/svm-bs.html index 0ad4b19d0..b7903aa11 100644 --- a/doc/pub/svm/html/svm-bs.html +++ b/doc/pub/svm/html/svm-bs.html @@ -174,7 +174,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Nov 22, 2019

+

Nov 23, 2019


diff --git a/doc/pub/svm/html/svm-reveal.html b/doc/pub/svm/html/svm-reveal.html index eaa2e0cef..a4efba160 100644 --- a/doc/pub/svm/html/svm-reveal.html +++ b/doc/pub/svm/html/svm-reveal.html @@ -148,7 +148,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

 
-

Nov 22, 2019

+

Nov 23, 2019


@@ -163,7 +163,7 @@ MathJax.Hub.Config({

A Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning model, capable of performing linear or nonlinear +Machine Learning method, capable of performing linear or nonlinear classification, regression, and even outlier detection. It is one of the most popular models in Machine Learning, and anyone interested in Machine Learning should have it in their toolbox. SVMs are @@ -171,7 +171,9 @@ particularly well suited for classification of complex but small-sized or medium-sized datasets.

-The case with two well-separated classes only can be understood in an intuitive way in terms of lines in a two-dimensional space separating the two classes (see figure below). +The case with two well-separated classes only can be understood in an +intuitive way in terms of lines in a two-dimensional space separating +the two classes (see figure below).

The basic mathematics behind the SVM is however less familiar to most of us. @@ -281,7 +283,9 @@ plt.show()

What is a hyperplane?

-The aim of the SVM algorithm is to find a hyperplane in an \( p \)-dimensional space, where \( p \) is the number of features that distinctly classifies the data points. +The aim of the SVM algorithm is to find a hyperplane in a +\( p \)-dimensional space, where \( p \) is the number of features that +distinctly classifies the data points.

In a \( p \)-dimensional space, a hyperplane is what we call an affine subspace of dimension of \( p-1 \). @@ -296,13 +300,15 @@ b+w_1x_1+w_2x_2=0, $$

 
+

where \( b \) is the intercept and \( w_1 \) and \( w_2 \) define the elements of a vector orthogonal to the line \( b+w_1x_1+w_2x_2=0 \). In two dimensions we define the vectors \( \boldsymbol{x} =[x1,x2] \) and \( \boldsymbol{w}=[w1,w2] \). -We can then rewrite the above equation as +We can then rewrite the above equation as +

 
$$ -\boldsymbol{w}^T\boldsymbol{x}+b=0. +\boldsymbol{x}^T\boldsymbol{w}+b=0. $$

 
diff --git a/doc/pub/svm/html/svm-solarized.html b/doc/pub/svm/html/svm-solarized.html index f9d5ef494..722543128 100644 --- a/doc/pub/svm/html/svm-solarized.html +++ b/doc/pub/svm/html/svm-solarized.html @@ -110,7 +110,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Nov 22, 2019

+

Nov 23, 2019












@@ -119,7 +119,7 @@ MathJax.Hub.Config({

A Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning model, capable of performing linear or nonlinear +Machine Learning method, capable of performing linear or nonlinear classification, regression, and even outlier detection. It is one of the most popular models in Machine Learning, and anyone interested in Machine Learning should have it in their toolbox. SVMs are @@ -127,7 +127,9 @@ particularly well suited for classification of complex but small-sized or medium-sized datasets.

-The case with two well-separated classes only can be understood in an intuitive way in terms of lines in a two-dimensional space separating the two classes (see figure below). +The case with two well-separated classes only can be understood in an +intuitive way in terms of lines in a two-dimensional space separating +the two classes (see figure below).

The basic mathematics behind the SVM is however less familiar to most of us. @@ -236,7 +238,9 @@ plt.show()

What is a hyperplane?

-The aim of the SVM algorithm is to find a hyperplane in an \( p \)-dimensional space, where \( p \) is the number of features that distinctly classifies the data points. +The aim of the SVM algorithm is to find a hyperplane in a +\( p \)-dimensional space, where \( p \) is the number of features that +distinctly classifies the data points.

In a \( p \)-dimensional space, a hyperplane is what we call an affine subspace of dimension of \( p-1 \). @@ -249,12 +253,14 @@ $$ b+w_1x_1+w_2x_2=0, $$ +

where \( b \) is the intercept and \( w_1 \) and \( w_2 \) define the elements of a vector orthogonal to the line \( b+w_1x_1+w_2x_2=0 \). In two dimensions we define the vectors \( \boldsymbol{x} =[x1,x2] \) and \( \boldsymbol{w}=[w1,w2] \). We can then rewrite the above equation as + $$ -\boldsymbol{w}^T\boldsymbol{x}+b=0. +\boldsymbol{x}^T\boldsymbol{w}+b=0. $$

diff --git a/doc/pub/svm/html/svm.html b/doc/pub/svm/html/svm.html index 804b3f318..8579530c4 100644 --- a/doc/pub/svm/html/svm.html +++ b/doc/pub/svm/html/svm.html @@ -115,7 +115,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Nov 22, 2019

+

Nov 23, 2019












@@ -124,7 +124,7 @@ MathJax.Hub.Config({

A Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning model, capable of performing linear or nonlinear +Machine Learning method, capable of performing linear or nonlinear classification, regression, and even outlier detection. It is one of the most popular models in Machine Learning, and anyone interested in Machine Learning should have it in their toolbox. SVMs are @@ -132,7 +132,9 @@ particularly well suited for classification of complex but small-sized or medium-sized datasets.

-The case with two well-separated classes only can be understood in an intuitive way in terms of lines in a two-dimensional space separating the two classes (see figure below). +The case with two well-separated classes only can be understood in an +intuitive way in terms of lines in a two-dimensional space separating +the two classes (see figure below).

The basic mathematics behind the SVM is however less familiar to most of us. @@ -241,7 +243,9 @@ plt.show()

What is a hyperplane?

-The aim of the SVM algorithm is to find a hyperplane in an \( p \)-dimensional space, where \( p \) is the number of features that distinctly classifies the data points. +The aim of the SVM algorithm is to find a hyperplane in a +\( p \)-dimensional space, where \( p \) is the number of features that +distinctly classifies the data points.

In a \( p \)-dimensional space, a hyperplane is what we call an affine subspace of dimension of \( p-1 \). @@ -254,12 +258,14 @@ $$ b+w_1x_1+w_2x_2=0, $$ +

where \( b \) is the intercept and \( w_1 \) and \( w_2 \) define the elements of a vector orthogonal to the line \( b+w_1x_1+w_2x_2=0 \). In two dimensions we define the vectors \( \boldsymbol{x} =[x1,x2] \) and \( \boldsymbol{w}=[w1,w2] \). We can then rewrite the above equation as + $$ -\boldsymbol{w}^T\boldsymbol{x}+b=0. +\boldsymbol{x}^T\boldsymbol{w}+b=0. $$

diff --git a/doc/pub/svm/ipynb/ipynb-svm-src.tar.gz b/doc/pub/svm/ipynb/ipynb-svm-src.tar.gz index 5ee1bf504f082ec7f3047a0f7141ca9e744333f0..b7c9245f69e958604ddcbbc74ac373644a9d3a35 100644 GIT binary patch literal 189 zcmV;u07CyCiwFSutJz%u1MSaC3c@fD2H>uHia9~ZqfOU>UAPcLyg*7*8?{MIQn9zU z573o@iy}h4&Cd+OFtg8B>urww-A9YT81p27F*BY{iRoNV2%$h&$~mKHmJz^&F-{}E z%boPnIWO#RN_~>rN&R+i7^|ued*)Z*nSbI?D+ilh?`xw$8!roE&fsnhQf;pbD2;Su rAFRA~%1RKr2T_*FE2G6FZml_5(KruHia9|@nkHQfcHu%0@d7DLZPX?;NyXmY zK0sFrE{X{GHa{~A!^}Qkt+zSycONYVW6aYO#>_aI5+`##AtV7p8D~7B86$uRrJO~8 zmpkdDb6(itl=>vKlltx4FjiF`_RO!qGylY)Rt`41-q%KhHeMFSoWb21q}pB=P#WpR rK3I9}l$9WK527rUS4N9V+*)(AqH!qx^(R3P1i?2R4kJu&00;m8xN%VV diff --git a/doc/pub/svm/ipynb/svm.ipynb b/doc/pub/svm/ipynb/svm.ipynb index e5c37dc32..181342b04 100644 --- a/doc/pub/svm/ipynb/svm.ipynb +++ b/doc/pub/svm/ipynb/svm.ipynb @@ -10,7 +10,7 @@ " \n", "**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n", "\n", - "Date: **Nov 22, 2019**\n", + "Date: **Nov 23, 2019**\n", "\n", "Copyright 1999-2019, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n", "\n", @@ -19,14 +19,16 @@ "## Support Vector Machines, overarching aims\n", "\n", "A Support Vector Machine (SVM) is a very powerful and versatile\n", - "Machine Learning model, capable of performing linear or nonlinear\n", + "Machine Learning method, capable of performing linear or nonlinear\n", "classification, regression, and even outlier detection. It is one of\n", "the most popular models in Machine Learning, and anyone interested in\n", "Machine Learning should have it in their toolbox. SVMs are\n", "particularly well suited for classification of complex but small-sized or\n", "medium-sized datasets. \n", "\n", - "The case with two well-separated classes only can be understood in an intuitive way in terms of lines in a two-dimensional space separating the two classes (see figure below). \n", + "The case with two well-separated classes only can be understood in an\n", + "intuitive way in terms of lines in a two-dimensional space separating\n", + "the two classes (see figure below).\n", "\n", "The basic mathematics behind the SVM is however less familiar to most of us. \n", "It relies on the definition of hyperplanes and the\n", @@ -138,7 +140,9 @@ "source": [ "## What is a hyperplane?\n", "\n", - "The aim of the SVM algorithm is to find a hyperplane in an $p$-dimensional space, where $p$ is the number of features that distinctly classifies the data points. \n", + "The aim of the SVM algorithm is to find a hyperplane in a\n", + "$p$-dimensional space, where $p$ is the number of features that\n", + "distinctly classifies the data points.\n", "\n", "In a $p$-dimensional space, a hyperplane is what we call an affine subspace of dimension of $p-1$.\n", "As an example, in two dimension, a hyperplane is simply as straight line while in three dimensions it is \n", @@ -171,7 +175,7 @@ "metadata": {}, "source": [ "$$\n", - "\\boldsymbol{w}^T\\boldsymbol{x}+b=0.\n", + "\\boldsymbol{x}^T\\boldsymbol{w}+b=0.\n", "$$" ] }, diff --git a/doc/pub/svm/pdf/svm-minted.pdf b/doc/pub/svm/pdf/svm-minted.pdf index 3c8036925cb93d3401c7d480b479dbabf274883b..3f1b4198826354331060b423cb7a2e09b719618f 100644 GIT binary patch delta 8891 zcmai)RZQFsfW%oS?oylrE$;4a#kFW*S&A2T=a0KXS=`;NxVsg1cbDQ6y)U`k^CdG6 zlg#@hzsyoLR#OgE+!r_wb~dQ^J{%oTsiTm%)`;DCs{MDk6vt?%2kN7q#22HWu71BF zvcwPJPa&fsb(HYav+?zL0!;rdk zyMsW{IVvzcUnX-b{@8k6*Xt}X7J>PMgxHMTXzlX0-UX>)?Q}DS!Lddm^fp@?pKfC)T1ud(d z{z&GP%N8mUkbO1&T_lL&GDjdHiA*e;jl?tcV5HJ`QmC^bjK$N>8<6 z61&T>>`8f)t$GP7NJuH^bGc#tsj<43^+sSRb+Bxe=#B?xeKF6M;-oJ z#AOyxm2zIijmGb6WmiE|UGUWlVa9w&hQsa4q8BN;_l`e&u+USRA0f=4j%`(Ch{esH zuvcN@-Yzl}TODlRfAog12nt@1isN)EZ07TX($a|J@!0}|?s)i<4}ovC@j~y>yJB9-r-=SDfXhtW9{VB9=s@-X07|>=)oz5r$3xXC-j;Y2Ti|K6sR( znXn%$@;hFE~^5so7jlDEOiO1BOe|2V4|za*|&;ox46n zB&@vHWfh1&yyT}#ZH@Zg9n43KMJZ~^Bgzp`yR4L}loQh!6u@^6}98^V!q_oOBNIPruz* z@0)Eb3?c?)JE!ELE#!mtw$=ou4vbP~!)8lJ=)3^o&lcHd1}BFI&?$*^dmg@X2<5)h zil7b_z6hJlBDJ0XM3UZetf9FoIAuFVWz8!&P%#dYO@sg=b#Z>ygPApj+?1B^&Jkc) z_7z<3uB)*@^AGArxBNRQq4I+@0u}mNv6$5u6&+S1#yrmxUf^)uuB*zD%ii}1!mB{C zMIV50C+FrwX5aEl`bp4Oo*OqiUX`k=>mo1T(!ZWqUiD{AF*vAdroTM3?uA3PJQae< zU)}(Eq@M1i`pa$@pHTlM#oB0LmM5y(&LBDn-vsX7tmGEQR(rt*sw#^7Kc8-xtE3-4 z@X~Tp3>Q0{W7)Q~{QPw^Hxj|9f38K9zMcjMd$h}c2K)IqTFocvTwjNDVOf+mD!5mz z%#*cr-GL;Xt|)7%-(yDXC`TZ+gQ*>vP3?5L5hZ&#vS_m`9O1>&@*Xm*vdNt0v?b{y?&`QtVz@xc(A-R@uQQ9B9g%?3I2A%2a0RGUE&pci$&Ik@}s0ba)V?34t zRNoF}q_~vG$nl^BUZJh$Jv}mj)(i;qBJp@~VE*nrLLb+7fVS=K?0;)|K}RZV^P%tg zP7~=BviLw?i*0^Ne*Y5HMr11Ax67s*qMqJ-7ejx%zu9=a|GtjfWe9_$jo1DHg_NH7 zV;KB(+oVDO3}LqB|Co;UC^ox29_9O%A*ws}ym;K1h@;RZ+=c~Q8Tkl7(g3iNt7qe1 znDJev4h7bqAER&6x2ocww?Pd>HCZq{z&J=EdgZ_7(?s0@ z8#3W{@=^$UME6|p>!N6$LUX&Q!R^X$ zK4d<>fO8fo_)%`Ih`+wm<;&`_>K$v9Z;bby&8%tk5%IWy;L&z|-W1uZX8zG7f5CT` z9xbtLSXH&8AFsIg-{~36+E*~eq+NYKr02o? z;Fs2$n@Pfib(=x_V`^Z7d`=$Y8JxaHMMlYH)%w-q<1^xnY3MO`C$6m zeC-sJ?9vf#aAAb8SAkr8&d>^cD$teF^E0ZE6TK^7sPecUdk~K!>X-;WkM33HZ%wu9b6F|R5<3z; z@#WuhD~j@;DYoP*>4w? zw6J32F4ITdKk$s57hCWP?VFWjN_U}b;Jnx*zVaF7Z}@ZZku?Y3Syhc9>U?_R3-YWdrINq4R%u8T@D1bG8Jg|Jv#9CF}h3 zAHP(Kp;ZvWZN+d|d#6)Zb7%DSl{j95-_EH*I7hxF7?9L#oa{J}lJYBlZUuSiG@h2& zC{RcXSE{|`wgp{SpltJhMvM!IURn@$34F`F&EOMfD(qp>N!suNy0kr=t5aO1ED~|D zmdkgoWpIr2<(*r#%T~*W1MLLzwWnosHYp<%;N2q)q5A?cx~^8RxyLch9-TK%!74SF zL`YIosu9?^c@ei~pMOPuJ1p1}8_1AH3{%Ik7K4Q46wPMECe84b$O8j)r)uQ zqHb}CrAc||(Itw3?SWXOhKc5q6NP^b-=Br!+MCuf0s0#^Fm-brwiy+z)-jZ=k2XaO_9H9N5^5~DbaE&e{ygD6 z*n{?{i4Y8ov9~zxLnH*@5W*b=#@(Up&gf_^I)8o4#Jq6~LM3JJL)dzpND|S$gwNUX z968T_%C@hlB*eJ;(I{<;Y8YsfWkiziEKGVr2Aj-mVN0N<^n|o^&bQz?$cPTh)o!P{E1|H@xJdFtULGaX{u8E zDH9ciJmk7r)fhXSKTIe$#&R2icsVW8C?hL7C?#;TE>%GRx>Z#Q;o6Md!TX|kngo&f zdx8U`s1Yoe3t)2gX&TM>GK5cS>PjcoudXDHbZi(jk`#stT0P>Ei~+ zaZ=r-)*>6hmFcEQuKpn0=gL(f7wNm1W)Pr0zYaB;%BO<_JAZ)n5 z8b(a4maPnL$BH!bxDKjiK?R{R6%E)QxKs*#Qhq2J#Y=!0ZL2=|w;#5aqJ_rEf>~^h zyHlKIvNDj9K=s+D1?OVg(hR~}%r!aLhg$fxgSBS~iZw}(4b>RWt(8v)DM2B`-S_}H zwT^;CwXYA$_i-jCk6^qI9m?kny82}-Mg>1VcmlsidZrc-6G)Y!-OKz;aNP_6j==KP zn*Hb(E^`rzj1qi#`>8z?y|?77uQw8|hPOq1C4e8*vu8G9B{r9_WG{y?Q3};feM;`D zMnCcT9bc(>icTR?Y-a#mmk+!-*A>u>9xMZbvEaXq>>srP!MwKmUL!Mb>!8_|-I*)i zdgq1?#(_Kw=_$OCek;J-E=_a7lQm7AU$l@ZJ;ErK{Gk78$Qkh^{I}&d1C>3U09=fp zzoPV?(kSgixGRQ~^cP0!!RU)K{lBX6AQb-ad3QG!_baK9sx`n$6sycWz7r5ZHWa0g z14kDZ^}?bteyr;sGnZL&9j`9T9(&N4ml~&JvT&q?&cvE~g+*;V?H2>Rp`e|Q%cQ<- z3+JcIT7*y{{|V3RZph5+&exU;As+Q3F%%`AmOr7mY0|ExN}-IYOEOc{hW~^?w$Xc2 z7KL;sfqFwt^b$AuY)OvItqX*Ogn|5^#aNknx$u{PdWS{_LUa-t#?_@DNG3CZKL?2T zQ8c|P+gPx6FvzAx1Rpl|LN2s8$_jVO_Rm1od7OWW9emvwij>AL+%S@l!z`x?!P~41 z7VW9`y5PQ!&8U#OHxRlMBI~!{L*r3DJL<|`*fq|M=dMc`bgV(zLOI=GlxC>eS)$FfXowq&4 z%CRz$Lw?kr^99M>D_8JzTubryy66aJAw41o)DUd3i6GxZVYMFOqvELDUz8xPB77&mQKu7PpoKz~F|6{DMQT~$$_``-rl z{3i)I-AB@V zh+@wur0~n4gRUvWp~c|&89zPCxMD$Efh76Q&Nid4m!Y5FFktVF_pVDC$fo^i^3r#N zT^ZTD;?yCrPa>lB+M4nkA4$g=!)b+80ah7p7R8dR0JR#LCh=ShdJe)R729}?aZp~H z%s1mi%X)$a%fCtj4I@##76D$!B8y>G?T}26s?^P=B}cvvKc@Ix<06>6GtJ_$o^e>* zhs5KJj2su*J3w4N<+KjQCW+^qx5wj5^O6Gv(oHZzw4qh!U?pYyu4ML~s=uxxJXhE0 zpoO4aYDL~)BU3$=Ea^bIJ++edpF{T=gVF9j3KfQk4n0M{@GY$^+6c80yfR1sX+(Ec z(QJ3Tnlf0U?+YCgK*L2=iNEZx{@GjFF;RO^qwKW?kUO7f5NwsG7}QYbJEBRD`~PDu z?iUbFvurL8T+E&U$V$ABgpZRmRzI~RlotXhf1Bu#7MTci*fHoX-Ly$ zS-AB9Pq!)=f+@l?AX$iMd2!!peAuCk`(sO)u>m$Xxv+ZXsz(j`g{!dZ3K#sc{`+P* z%MWV^t=XGj3WI=>D(L1X+ga;9H~CshbIdu7W1G(rVdc)p)u#gZR5Y$tKbSvCi^ zLDz$yX-BG#I%RZ<9MW3M_%`>W_E8>bLaAv2vwqyu;TnQfk(|!=-b#E_4xA zK0VR;QsFxzeR3jKH}l!_CWD_9ZF79}I6@kyfvw2RTF^aHhCgFVy8an07(Xd+`MB$) zXu_v#J;kjZ*~_U9mfT=t=iP^YhL%SXMsaLpSP!_`n|F3%5>Pz9*Tc9p=^W_V4feb=H|H~Gk^XVXA#fd33kxjCKSe(D6mE~v4_|Goae&mrtA~3L$9^t?@d5P2u+r{y=^mOKcA*M*coCfEj+Y&qrm8eyWVE!bP)n z;s+UpkvgT^6~+@ZiHc-7GlltxdxVUw`bUh3_SV?D=QBrB#cUkNf6xpyrw-v2D-2Fp z=bzu*$6Lh9hZ;QcDP0)ppdk5v+(?G4L`9++&Fq>HSPRa7*K%0lxr49(CAA;Q2MU^; zwDT-V{Rs&y)sQWma8LoSHOLX<#f>Lz0u|Q|I3Q%g5jIKj>oh}s#Iv4*n9%W`sj*G- zmHC+!hFuRJ#i4W2$vB3@P|KtIg#E_Z_IZoVdi%-LT|BsKn=6jO+_r_p&ps+GCbJenR5hnkuRJ_H=Q~OSI(K|B=ehNuYEdp^u2X#Bm~2~(W)Tc{k%6~ zuUKf%WLen2i<*B(jN%3orR(KZOpHefi{g8^qYzkZ5O;*NgrBf9EnEbynyrg3S3l#I zwEy`o^&xvVnTw6y{1a@~4ez$weO8DA;@DE&fW6+-16Mdci?A%!ZkK4%=x|(q!K*HKMDy^zpc8r|Ael>fS7rYp4S7mLk$+tY*Tn*uwFT z2abNc=sT`s>CrBp!=CN=sOTd2C?x;Os$-fp)opyN!;?vz&2%{1{TgZ{XW zI_nh<^gv3iU8S<2{6Yjdq?s0wyf>3b+CcWW3=!bt{pL*Y!sxJ$d*eR#&~MpEt2I8K ze{;kC*lspD>9WR_qTKp0_x5t_?n<9=OLC{^Q~qfezVv`Z*(Zpa#85`6JMR2wKWuNX zcdzLYo1bGCdXx>~!M8hU3miKwPh{u1m@-a>0Y?GXN|6<0=q`XLk-zu zJH4w)9=$_^!M;Yb^P2>;_5b#vB)K4J;JV5lrE(zg!7kAvm0mz}Wb#ZFnb#Ru{4HsY z*ugK%rNOnb&V>Tj-xyCEVJ%Oyg7p)N;-z z>f038d?B4pZ%M09oXmo3cdd?a<=%*t)(Th=pL}j-!n6#FKGeVwKX%SR!O*JplsLhcREx`$P$vod*&I@aG%H=DHoCiX90hk?rs&lBSzDAae$9Pqi(NgGL6C zI@lZew~|zJxsSo*>rx=Df$4r3|Bq*@=pWaQ&`#hH$>>}X(KrTh8n$$8rsL5arH^!R zh9e3abD(KoN0*p7;~J1q*gZZ6fWj(fW}sFk-k%+@H{8MjwwS~t38yWh7qh1(auGBw&jg)85{;ior5rus}=3t5cY1k@d%E zpON`GHu+k6*Th4#+N_QKKJT&l8>lNk4zwM#h92dE6oE^Fb1n6;%RPyfz8+3Y3p#mS zdXsVkqzi&&Scj1xEIUdz1;u-&BkK4vXq8i7r+Ik#@^o434a(Wp=#Qu4Y+h`1bAvim zS>X$2%_wMP@;k@MpH(1==c>v2#R;-R<z6-qJ7f=AsFc?`c^-XgaMK)!;d&WV0L|w)zBsfO10^ikW$1hX=o|31B6Yu) zx)w=`DuKvcvvbuEbZ1JVC5594Yf=K{#61Pq{npP{ua_A_8ScmTG@&JxR`9_qPHs`p zR!__?Uk)#TEeH$@^#2Q$(EW_%9+r)<628GM``n##`eio!pe=E@&eQn|vK&XAaFVdO znG^>x7F$=XDgvBKp1V~?NFxxM^h<@g{+E6pW8eXTmGCbLSTl%H*hpt#+AX18jqLL5ia%zx|EIpc02ycazeUHE-MLM6UUj21A>?yF)RrxYGF zu&T5%p?n`A8a+R{v&+h@;GIH1c)d){ZE*BZuT9n37P{+PV3ftAXUx_sQtx+MD4bLT zzAMZm97I~(3y~B=)iC5q=9Ph6Ak-$SIPfV;drVJb!Poc3==T2lIce^0 z4t=gQNeEoUiNbG9;{bqdF{y03pZ@KNY)}lh(Cf*iItt$QkM|MZ5SDRbW^oi7k8qX# zJVLZ0lA6P0^~!1)c2HV*_Dt$C12CkOqi)@FV=4zckP&t9 zlyx(3PLxu>Cu(6np`)eZ;*Ut5a3fgd;ZX9g~^TRDwR|!BKDw zE=eF3^q>SJj?YW%+wacNCeSL>p)<~U`ZbyyW8>x&DqIR8{#bdSfE&X4uS+S4JE6Q| zStGV8=7a!hRSMz(`9kwbK{UW-jq7IQ+q_ox%ilw&%BT13=R?+mOU#MrH%GwNwcTe( zzUr&tpzBw54@d8?$cVnrZ zsC3FA;EeH!`O+u5JLSyMhZFrj^TDm(w14JSfu@F(Ohygj{*B#`X9)o-hLpyl!HS`kxv-Yrc94wJ=tBU~kCt?0A1wzD0O26vP5AH}ylJa5IO9 zk%`r!OlF7LDd_SKKkkA1r?V|b!PW(h^T$oV8?s`zxaGNSTrOymzaSVFEnG8k+x-dr ziL@h~dBUuV2}5Ob!+qV`@Yb##^L2IS$3Ge**No&#USfQbYAdsA6kCE99L0%e$b6Tx zUFaAKQ3G&O2~Z!;bI%3jiHLNx@(DgR+jhT1Dq&Z5ioUZl8Ndp%hRT$K{xfWoOgS8& zg~-Xl$;QUP2^|N0ApO74bP4)kimEQdE5XIj%f&0f&dDVuCBY#jA;~As#>UUb$;mCw z!NVy+A@u)N3{Y4VO>NDcEhxBn`6>SQxIi|;uacyOH+Xfk<`!b4BZUoB!yoVw6Nh&g z@(~wz5MNWBp6SPgm>L;LJK&PwcKLCKhp0frjjgTexEdGIQjT7-51@w=Etk((f`>9v zFACW71#XPuI9&93ePbV_1J~fKj7Xd%MXmi@&KSj0_dSmS&tDc$K1y@H%$Sv^1C`>K z*1^gs%mqp{@%zLU8qPlV_2)Gx0ub(JPq~e?(xc}KALq0q{uXXxdFM0>p?SQ!h=;4( z1|mMKyy&@nLNd2`y8}{7)H8qIfyaV!kyVT(u^vN`pbmSi z5IoMY4Gy&DZJ4N@uphqT^J;t_fUjFjjpGDhVUmP_TNTMoXdU9ABQwpv&F`5Kx7ljh3I48A@tKQ*tsR!Y&mzjzZh z^S5Ah!YYb#zK+9>NOo^={)a+DoS3h}rbw!?RMs@f{Wke~`g=FKO;*%p`hIZza}1G3ZrF0?qw6+*g--Wj z@J#s0yJ-l+A`* z?)2&gAihaj)mElL=|T9aX_L&hs?a4L^brp518mt8=@0*e!9M%Z4-rVPA*`>GO0{*- z(CxKR)|#YHm9BWX7Pj_+t$?0?hVzNxyBqr$RaDi=V3?t-k`fqd5;W7+!bt__M624^ ztFy{HY2Zo~8N@1G>Le>vlec0Ge{|D)fD4#5McvsV9e3RRWsgL#P#5o*^0y&;L7Jap zm}&BQ#h+t8-W0IXN;g%KHA>t1`X>Ju`Ix{EvL<~W=V=_q!^Si$qFLvEiu9A4!wAS@ zG9YJ@(T%!s*e=}+IOCZQ&s{&_HoWX7_W+CyAj|=%o;OXI-ZA6H(BV1z@aP8Fbau19 zsWBAi7W7_2QP&qxG+L{sK+S+^>Vj4CX~g}$z&V9YjX`OOA~m^g1w`PBTD4_q1SAMV zn`Bj`bX)(Aj&sM<5eZ*^@Z%>pt<#U<$>9)LgsV$vl^eKRo`76%kD^VisT6opa(@6( z-fuFSYmvy*ERoF(NYt2BBILmM0P`FlknlM#^PhkD2#sa}h7?@sc=!d`J=ly!vOChA zcw)EoJVh9}SP=m#y?=cwhM`=w5Piat5%~Ga+hmp{;_5B?aI(Kvr zxH%38(a;`=2_zb)!)a*`N4iNEr5(oMlHBz!#_%~WY)9ypGbaBy`IxQpz7jWw096|n zH_9?CYAAA>dWP=6qj8!0;3pnz{R}37HGmEU8Uob+Y33;k=sCD4#cy1=<96N);V?$w z3FK12nG6in1k~B2thYnzJ2VpbGC*w7Dt)dRJz^yESVVsPU3$&S(6v zpg;>mn>{FP_S+MU+AGu{;ljQ!iu-ZtWy#(lCu2x9nXfkkO@!%9@#{vFm6&B78Ls`q z8Ma?R;Z@zFfO!IXow~_$2ZLqkP@6r$y+MovqWYdkaS3#PGvwW|L)EX)!&I~yF6qX@ zvG-q@_^`|@_Dh)|s8qSAJRV%%XEFSTbO$U^Wv$Ex| zW!OR}SPemcU)d%M%;APNZE20FTG5NZ3rv7Hp&@XTBBx_C5`j=sS_I-DGKpNM65+l;-$egv`{LX4Nf~a$nuw`y&dRpo*w|=@ z$Phv!_o=dIauE~=TaRiR0{bAb?6}=i<*;UlBiyrh!@3WLv_`+ZQGVLUNojCd=dLD( z7!H+xW_x{ugPR`1a3h<)$R_{^(^@D_#Bxu2sfw@+$`NnqfB-p?8YpM%zyPsBsK*KN zZR~NKr->D9eHEgR;W&+uLJyHM|9MFc!MCj7xPU1(N{RxQlNn$-$hoeMw_$D?qgr#a z;u@}YYb4{$8d(w;uac-&xmI9p>KvvQyKQYeTyH!mF- zPZtz@powGFLEI-*(QpSZB@Qq!#Bg))GkF>e$)hw!hL!>$4B*<`0-ykU(I)mq;GjBx z!=}Gt?R{Qy($so~;^if5t@5(uTad-UWUqAEuuhAjxhS|qlh}cZE4!LFg3YBCIYX=f z%|hQZk0_Vl}{{>7!hGc(M4H}fm7xnrP>VEC{;1fd_@`sBhb=E;S} zJ-^&;dYLKUJt&r;)TVAdU2=NtH;p{@e|Y963k1S73SRy`a9HbLhc5V=&k&1$EFj=j z^<~PzOt|;>&ZEB61Cm@Ye_p*(kiv;X-c2w*to0(M={0A|ON9qu9`C?OCJA4EnA3$1 z(w&!+#^EH+TPa6ZR<+u!0eel4nD(tMAsfVQ3P-8zG#3-tUU)1H&q<~}=@vj;=UD5` z%(FyUI=e|oQh{kIL$Fyr3d@jHG$CB)ZKRL-@^TBsEG+=N^aJ^4F&R1n}jgeT60ZOw=L{{E|6K5i9486CdBj1W2qJZUSAy+zBfX;pvGePa;{7< zYniqdAq^z`wJIU<>_R_IsB0RH%tFBc37qw?ht{Df^cxCH!iz%6&x(#yp)6(J#0O3e zNiQ-LA|sfPT@$5{gDFE?9(=0QZaJtn5dW>13GsTW#!jXE@S@sUqI>jzry_}3nX~fr z_A}erQuVK{{)c??eUm4(YHQ3oJLut~A}d<@kw)I%u}edgKeJ^}xIn1La<%^a`kd=h z7%%sOpkmt7&o${;wk?#wKa3pfSg1a9c;=N>{&4^9zoH_GHk0lM zCj&S%Ig`;!DSuntkK4Erf6rgxM=x&8<2R{o)8Y#B+7#$h7wrK_9%Ng)R^gR>mNuLG z_cudQA5OAG}nv!hXo^) zY4}DLY{8@@hVg|~hH#-4_fL!W%m3V1zASE5f@jOTu796z*zzg7?ut`*YIb>pPj-2D zKKZ6?-sHZiE*jso$t(Y`zFBe84*GHfG{t`J-|FI+pYo>I`o3j#1C-4nOoOB^*b*AH zz|G+)FRS|Y=HvZ8a5cPLj1WY!g_eqN>zDEp7?x_8`=7cw<>lkf&EZ@2Wfe%aFb8s0 z``2a{+JFChSo?>vioR&z9oTW~m(`UuRehg<5wf7$Fonm}LF|LE zX@Su3gMJa zA%EgNt8MvIRCRf%^1Yw7KIWU^_A8q}`4~lL#)l^UHG|v|OL0o4q!de*fGM;DWiS9Q zRCsXP@c9Xav-Ywo5kmFwP z&%t#-SFeXd>222=APO%NBbTMLAWQ`5y?;Yh2@NWp))dnj5!%EZCsftXW+V|Ni#v^U zB~+`cz=LZ_o0SN=1n6FZxRm#A;p-;fhcWYZO&~PgcCMe-ZZ7``^pNm@=rxpLgxO#s z9lO{o3%#bKl!1w5Djpz6Nw7f&Lj*Og^?aa|hCxLfYFfrPqr^%bkINF$kNPN*aeou{ zLEmZk@&_kH8>Eyii_iJfaqn~>x!=0*we`V1d;pmj1sEPkBhcpliwg-(UuLK;8}tQn zs)Px?ObT3Ng!5n^x$DCFU}IgdftC<#;OS*o?!%+!mcfoV6``q9Ss1BoJV_bebN@3~ z-wGpRY?r(=ckFig=|&ldYGuBLSTyH^!MKbX&HOoZ&_%1OZK8cp#B_Rs~-j#b z0Xrt{v_3PB2H2r_+y`xwXMe$p@d3j>fdAsODUOYAn%{IBwg#e^nEz`GZHi&84KY^5 zBcEZ$YFOii%p}cc(gRS4qHM%WyQ3o+0<_u}Pay>e;ziizMCl1WxioCkAm(?UPtES| z=z_;YkdxpK6c3&UQMD8y;gP~%Y;%Zi#$Z4&l*@RW^*nGuK*E)Z`+s~lCw(djW1|kg z37<2BHVDtfHLC)Ap|e-5>Q{D2!&Pa}F95ZON$^losn$bHVm&LsBaciBu%6MfyF-px%NX`AD zD_g*CSGG8dh@cU&f7l3pX;e8iQw`zn$z^5KUNB%wNb+Pn3p-q2bEFyMx%L`{1T!Q9M?dT`s38IeFRdTN!fYNqYql&dso2sk=GVzGhX74W zt-2CT!U;ldqR`r$cbWEuknY-R@?EScr2rNLM*6>-t4^^UZNrFd*mumw_V!8pr$nF^ zfS=995q~WeT)!5`F)0vDRZ1Y)tU!**5Q>QEHOc$k`B3F7oGXBL9zR|-7r$&SXV-V$sDJMUZzjnH=U|=Ln2~Ijk(AB#<{OAU zo^#QP@K_~>U`YKj!c0&wjDY8i`d7)*1I$wUZJ&~%g;>Go4zQv#HBMBfQaEOsEB#G%)yoO`8`FO8J3lM7B`2z zyIfFp97FbwXN8Slb~x%4bnf%P((JzK|kBUaSPx z-o-V$vP_0X3J85b)FRY)N}?E;G;_Ic?|*9W03Lvp5BBg(m(7#FOdbVw#f0gvb0=6~)el4aE3FP%?*ne1z zz=MIHly?tN-9-Aj3D)qq8PrV#xPxk2_Jx!7MnZLHe7~n6TrYTNZ>i+|!(-`(2_46t zZxmNG;nZFjSZ914z#G7y1EQBf38+h5Y_cA{VgP*v=C#8D&}$FY_=Twd8ff&5w|6=% zkl*IO@EREQjs3_a>-r6*JMxlr=YKkhrN+WaG?w#vvOlHk)rxfmJbJ$Po5UUcmBs)K zm<1mMLDmBg^g-OTcNJ}@ww)=rP(T*ZFt0CA4LRgg5sIg{!=YU&D5h@2^nE#X)vc&6 z@kbl3iiF2~yE2p);boJ%VqJSpBs|x-yY1p%+`X;o>`wsXF(YoTLOzp=5q}aOz8Vd^ zFI_DNl@-==r~=oAhnC?Y+%qk=?mlIp8+=ZIMRx&=hg%F2*Rw5_)S=LM1R7d2tFC>j(-3F{P!aGkbl&G!}%eZ!RZo)MD%hR5_W8{ z?!+x(-3holgLXg=v>7s(i#U=@v zHRuNv4)g_31gsA%eehwg0IHA+Fb8x1xd4)g#T4*K5KzImgw0{gK9eth2_AR^jei9~ zN&poIKukb2yFH@=;IR}SaKkE50r1|iB?D=!-2l1}*+rQosEjdSm?9Y_pieyl(@+3} zL3b!Q?imO)8max;%&rJHEL#%RFT)EeR3#u^963Wd zOEGNWa$Ugm&`Z=H5GWdzkF!8Yu^aS4<>LY)0icRnLvFGlm#BiABtjny71WlJ_LoAQ z0a<^b!z62**Yo}SxBucMM?eE*3nPF&zdk#Avy0LRC7>M*bYTK1)M!)i@iCF8fKe-k32Ny9@FjtR8;I zDzrMicwDyA`E8L4bl&DIGAeX_OFo|A;{o_Pe-nRa?`au#FpVv046g-+Pm^fj!$-uS zNwij{F(dv=kyR(49`IcP+7RMe&b{2$@QeoHc4B|NCw!sNwm(8y_pcEcS7E~p$^Us72O5@ zN?iBEJl^s>ahmUm6?66I{Co~RubDIRa~(IPlkPpWKfgSlT?!}tMmgUhQuu~b zA7RPHml2s`$aeH{Msg!F&QSd563LfHpaC`(PwKma%j>I?FTS>4Wc@{P-(8=auO63_ zpaCZVbC7G)_)1{KPcx8BSuk9pSQy>+S?<$sE znZ+b)>7(oeqh`K!1kUb$m}fs3^G(u*IkY?Y{Qv4*w&D00yh|~+FfI3sia9?hD%uQe zaXxS%jj<%&c8!s@Uonrqe6FbYe1<(ZbYiblJeM!}oQdszjqMIU4qA&syj^L`r|hGa3* z>*u?f1+3z?+2MG#I5__W6OXgwYxO^e^N+{s=pP>r=ks@J(-?TSJJlL)&^B|3deqKO zBlbFu0WSWoIl~lx1={XByW7O=2j+Y0Pr0o>)3O6o?<+U3%6AaLwgjQ*(MLax*uo>Y z;6Gbrn~*DaSGqg$hN(gBb%c??bY3TGM+im>JQl7fmJ!zanwR2kXr541 zv&=D}Z1j!yY7UvPmRYbgMjjgM0&HtCq5gd8NtUAlJ%8Weg*M(>$1$GDVGJd_SvXTh z-^zedclqnb=HeTNbuD5=c#hH!k)VX)7ru291<({n6B`9qux~e1bfYt<7aVk;)J-zT z6id*8XSGE~HpmsBXGtI{mfC_|Y91+Lv-m>NA}GY9rvBNEhwVIC?}eJDqrVVkZ5}Z? zLXfGTF<4T_E=lJi1xoNr8yzj*_s7=X+GK3wUJm_N=Y7}y$*!iCH}Yg{mq?@mOn>fO z?N(BoLZjT_;F$+B6s4>bvehmG_lsVlNjO{56# zkZIkfV=NNSL~Ao#1qM|m@pLDVCoraVC`+zWNu-R`M*P2wZarjPqVMf; ziGFYUn^s=OuMI%d7HSs|BGQtVW+F)Oj(hNh5P5OYyy2*UXqkgVEl!f(i3l|gB(USe zk2c2Imo}vVRDXVMB75>o?!}V@p@SA!@%UcC`Z5zMB8e;kGE}dbiFGg8TCP#4StDZw zWMT_FvJ;M?r9O&N%Ur2XsmamR>_k!l(!s&OKrnO45`n-&{B+Px^o~K=2L@dqnNBp{ zokW(^qZ>&`k`e8p>^UfC~ou$nyB@zy^ z(GW#zdN5kO#yFwbfT|)}6VtBl#FO+i4eZ*Evv*p8l)06bGj#No#zX5NZ=`N4-EtOs zirfV{DS!JMc5Pt}=~3=lYi%P7`9IE3Zbz?w^9*IZfi^Jcf42;!mCKWg&9n;VsrR*B zL3&78`P4F%B9^l*eP;UGT9TGDUtF{2*&ttv)C!I?*kv1}#ZrqBawM#yA zTt8zF-+^(Bd!>^Sr#m@GcTG{4_b{}26{gT^fPVzWz9u44W3+U}cfAc>wt1co>T_N# zYfV}E`Ob&Y+spSUsq3j5sC6lMyj*7#a@b7b{=D2CNZf$+^Xop|Ajr9ipS;>A2Td!s z!owQ#7^*W5o-(In$ux-{j_Nj+5t25*IB$$x25kLtj()I~K}7xVvgetL9M`11uSYp- z=6_J+#7KXa<=9%BmvE3p(Imd-+Fs2l%54YFo!oEY`wD)@DkNI6$jEI`M>!#($x+;d zmSbD;;+*p8+-Y?3!k^eiiK6U=Y@?M}x-K`7^OZ@JffhL%#(gHUi8ANMwc?Xf%RB2` ze?7PHyvcTYI)2(}xRInK)f2}Us#lQP0ve`BUIfh)gK;jVP*rjjaig#2nVK?$ddBC6 zmmsJCr2(#&#;5^CJZE28ROB_O(26VI!Nki=p%p=rS`vNv-wI6jMK2q@KmB30jhC#vHxSabm;&^_NGf0YMf!8otDMGP^puJiS=VFY8N<9}eF0?U$aZ z0U8vp=1UWXFYA9MFp{gi*;9|Mj_Mni&8Y!40lt?6ssSc{Z&-CQP&K*Bi&(xpPbM+(k z{Yw26dp}Tz>PQ`{nfm2Sy#*8}A1_X3=jv4bw|b|})O&TV=ITOSsw=fn*8u3_CNkwN z$o%E`4^JL{?XQrrpF$=P8UBZo`af7)>X6Adw82k9Mwd_M!amS>qP|v7p%u^6cj|lK z^pkoiIQ?hC>1cj7KS%lg{XxY{{epj|o51z3{i=Swo?XdX;$N@li`gaTUV-@!(YKq* zR=;AN3hCn?|NQLn&w%v#!Q$lW4~yCP6=s9dy_mg!y*@j*+>WmGRJ#aM^MPrf-F+oHbg-|G%__ZHaRmzJ|H|qH8Mg%GDbo{FgZj;HAF=+ zL^3osL_t9`GBq+bIWt8*T?#KuWo~D5XdpB=Ga!@U6exck(#vm6Q547V_1ovRy=^b0 z>QPmt9@U~Pu4iezAN6c$)mw>?L}Fr4111t9lRXjhV+V9ST`OpFj*te48&8>>A8SkUhM9;~s z(NLn(MB}Ml(fi8#_)+Qrj+JO5bte-Z}tGrz*&}zBAp|7JYz5p=XvHa zUVsUhgo_kB&?&eCm*EOrg==sfZYZ9;cP}>;Z~C34rH&6yx1^_Uoo-97r${+eN)Kx{ z`{QIIFC~2)cS_Q??@opE=ebi!>9IeR>y+LNI@PNv`J=f(#ofP7Wfh+%oElZk%Be}k z+