From 239a7c172c2ec28d0b4ddbb5c7536d069633c916 Mon Sep 17 00:00:00 2001 From: mhjensen Date: Tue, 8 May 2018 08:38:31 -0400 Subject: [PATCH] updated typos on project 2017 --- .../2017/Project/html/._Project-bs000.html | 22 +++++++++--------- .../2017/Project/html/Project-bs.html | 22 +++++++++--------- doc/Projects/2017/Project/html/Project.html | 22 +++++++++--------- .../Project/ipynb/ipynb-Project-src.tar.gz | Bin 211 -> 210 bytes doc/Projects/2017/Project/pdf/Project.p.tex | 21 ++++++++--------- doc/Projects/2017/Project/pdf/Project.pdf | Bin 195510 -> 195601 bytes doc/Projects/2017/Project/pdf/Project.tex | 21 ++++++++--------- doc/src/Projects/2017/Project/Project.do.txt | 22 +++++++++--------- 8 files changed, 64 insertions(+), 66 deletions(-) diff --git a/doc/Projects/2017/Project/html/._Project-bs000.html b/doc/Projects/2017/Project/html/._Project-bs000.html index 4e70bbda1..6c876cead 100644 --- a/doc/Projects/2017/Project/html/._Project-bs000.html +++ b/doc/Projects/2017/Project/html/._Project-bs000.html @@ -157,8 +157,7 @@ MathJax.Hub.Config({

The aim of this project is to use an already developed Monte Carlo program for the one-dimensional and two-dimensional Ising model, in order to produce the spin configurations for a series of energies \( E_i \) (10000 in total) for a system of \( L=40 \) spins in one dimension and \( L=40\times 40 \) in two dimensions at three different temperatures. -In its simplest form -the energy of the Ising model is expressed as, without an externally applied magnetic field, +In its simplest form the energy of the Ising model is expressed as, without an externally applied magnetic field, $$ E=-J\sum_{< kl >}^{N}s_ks_l $$ @@ -169,10 +168,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol \( < kl> \) indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz \( J> 0 \). We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values \( -1 \) and \( +1 \) only.

-We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages.

With the abovementioned configurations we will determine, using first various @@ -180,16 +179,16 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a random forest +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a random forest algorithm and deep neural networks.

You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of FYS-STK4155.

@@ -235,7 +234,8 @@ You can use scikit-learn to perform these analyses. See below for instruction on

Part d): Classifying the Ising model phase using neural networks

-We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12. +

+We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12.

You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow. diff --git a/doc/Projects/2017/Project/html/Project-bs.html b/doc/Projects/2017/Project/html/Project-bs.html index 4e70bbda1..6c876cead 100644 --- a/doc/Projects/2017/Project/html/Project-bs.html +++ b/doc/Projects/2017/Project/html/Project-bs.html @@ -157,8 +157,7 @@ MathJax.Hub.Config({

The aim of this project is to use an already developed Monte Carlo program for the one-dimensional and two-dimensional Ising model, in order to produce the spin configurations for a series of energies \( E_i \) (10000 in total) for a system of \( L=40 \) spins in one dimension and \( L=40\times 40 \) in two dimensions at three different temperatures. -In its simplest form -the energy of the Ising model is expressed as, without an externally applied magnetic field, +In its simplest form the energy of the Ising model is expressed as, without an externally applied magnetic field, $$ E=-J\sum_{< kl >}^{N}s_ks_l $$ @@ -169,10 +168,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol \( < kl> \) indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz \( J> 0 \). We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values \( -1 \) and \( +1 \) only.

-We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages.

With the abovementioned configurations we will determine, using first various @@ -180,16 +179,16 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a random forest +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a random forest algorithm and deep neural networks.

You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of FYS-STK4155.

@@ -235,7 +234,8 @@ You can use scikit-learn to perform these analyses. See below for instruction on

Part d): Classifying the Ising model phase using neural networks

-We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12. +

+We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12.

You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow. diff --git a/doc/Projects/2017/Project/html/Project.html b/doc/Projects/2017/Project/html/Project.html index 349676896..24ee5ff74 100644 --- a/doc/Projects/2017/Project/html/Project.html +++ b/doc/Projects/2017/Project/html/Project.html @@ -114,8 +114,7 @@ MathJax.Hub.Config({

The aim of this project is to use an already developed Monte Carlo program for the one-dimensional and two-dimensional Ising model, in order to produce the spin configurations for a series of energies \( E_i \) (10000 in total) for a system of \( L=40 \) spins in one dimension and \( L=40\times 40 \) in two dimensions at three different temperatures. -In its simplest form -the energy of the Ising model is expressed as, without an externally applied magnetic field, +In its simplest form the energy of the Ising model is expressed as, without an externally applied magnetic field, $$ E=-J\sum_{< kl >}^{N}s_ks_l $$ @@ -126,10 +125,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol \( < kl> \) indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz \( J> 0 \). We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values \( -1 \) and \( +1 \) only.

-We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of Mehta et al, arXiv 1803.08823. This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in this article is accompanied by its jupyter notebook. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages.

With the abovementioned configurations we will determine, using first various @@ -137,16 +136,16 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a random forest +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a random forest algorithm and deep neural networks.

You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of FYS-STK4155.

@@ -192,7 +191,8 @@ You can use scikit-learn to perform these analyses. See below for instruction on

Part d): Classifying the Ising model phase using neural networks

-We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12. +

+We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta et al.). The method is described in notebook 12.

You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow. diff --git a/doc/Projects/2017/Project/ipynb/ipynb-Project-src.tar.gz b/doc/Projects/2017/Project/ipynb/ipynb-Project-src.tar.gz index 81f14bbbcb2548af5671a665e2047aad8ff22fcb..220baf1c244e244697df723b159c56b028d3e5ee 100644 GIT binary patch literal 210 zcmb2|=3p?I^)ZTp`R)0$Sx$xmZHd?8j>a5Rw5(sp8e0-L;gWQR?MC%1;n^-PcK-gw zT)NbH;-lX=c5RYvGq?XN`DUg6$zSenZ_ui>*_O|itJ4;h zgql9}xboL;*Nx>lj~-5`%dOq?@70~eoUOYyKL6A8=c)b2d%IQiyA{75?B>6z=X~RY zQrp(QCt7dw)@jCWczI)T#{TbmQHFH_mrM0_KR8>qr~9Gtto0#1@_%_5kbzBq{J(JK J-wYZI3;=eOW&QvF literal 211 zcmV;^04)C>iwFSEl<-;r1MSbxY63A3$8pa-#XP~1?5?|KK@U9?r1S-1H@Z=q*d#*t z?du}^{N}s_ks_l \] @@ -171,10 +170,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol $$ indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz $J> 0$. We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values $-1$ and $+1$ only. -We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of \href{{https://arxiv.org/abs/1803.08823}}{Mehta et al, arXiv 1803.08823}. This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in \href{{https://physics.bu.edu/~pankajm/MLnotebooks.html}}{this article is accompanied by its jupyter notebook}. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of \href{{https://arxiv.org/abs/1803.08823}}{Mehta et al, arXiv 1803.08823}. This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in \href{{https://physics.bu.edu/~pankajm/MLnotebooks.html}}{this article is accompanied by its jupyter notebook}. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. @@ -183,15 +182,15 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a \textbf{random forest} +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a \textbf{random forest} algorithm and deep neural networks. You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of \href{{https://compphysics.github.io/MachineLearning/doc/web/course.html}}{FYS-STK4155}. @@ -231,7 +230,7 @@ Use described section 7 and in Notebook 6 of \href{{https://physics.bu.edu/~pank You can use scikit-learn to perform these analyses. See below for instruction on how to install scikit-learn. \paragraph{Part d): Classifying the Ising model phase using neural networks.} -We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta \emph{et al.}). The method is described in \href{{https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html}}{notebook 12}. +We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta \emph{et al.}). The method is described in \href{{https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html}}{notebook 12}. You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow. diff --git a/doc/Projects/2017/Project/pdf/Project.pdf b/doc/Projects/2017/Project/pdf/Project.pdf index aab4cc230e055990761a30b7acc440e65b087c36..e5ece03790ed50eca9e0d97ccbe1e381dc0456bb 100644 GIT binary patch delta 14535 zcmai*Lv$q!45(|{wms!-dutohscqZ3?bLQVwQbvOr}osg?e{;e-s0_&bCOjyIr+Y{ zJR>H&AjVTcaB!tL5rfl#QnYkj_FJ(1XKH6JFa??oA)r{tsM2M4($?&{rl_u{vTaG* zFl>KJM#-wWxAgXG_k+`^CS16Cj6n23-t6f1y5QB4sNbHw&+Os)e|tZ#O?fcrkL%K< zpDk^E%HpV+IyFqpt6Bc4x@xK1E+Bosvd$hm@#w_=F}6@<^ZMKZ+R)wYud!xk{V<&D zss0yM3EJ!yxYz%{^HRRHv9M-Tzk`m!g<@Zzt24>Gn58M34RC z+k!GO=Yq4L9F`H?hpw-nU+4OMn{G_tZ<84xpCb(tpL9W`5!fFz$sV6b6j<88$5b4=; zLdA*RijZD81w;Bp*G+0K&-S=2O^Yae3WGE&x>EvMXDfBKjDvz}k+Ifnf0YxxJc4s# ztGtQI*b<+dlM%N2^h3owpn}$E{=zS7q*Z9A_Q0!s`#F0B6hQ20ck$wJt|nFZ-ecqW zVE!l8daOyUshR6XSZ1{uMf7xfoPw4EiW{%G9>3?&af|3;h%^#ruT?}~{WC2Icay?R zOGs5wxY3^JZ|7oDroD|SP9#AV$4?mA&C9V!M`*n(Nun99Kl9jGoQPI>HgZFq_#_)N}~Z8g|W{IEb%gKtX_f zeV_MO@AK%%J7hi{3*o3+l-32hu+FVIvCNuPpE0l&bf^!j*V`-rjm=H{hI|)lT@7lu z4>B@n2L5v$dU*%Tp@wbg#QN+Uhy=|$)ccuza+K2~R55 zJMEuv&=onlhfnnBlNAPXZs;6$SMa()Ppv3Wu}CZDJIF)Q<%>pkAqO+3r-&#~iqBsk zfnhR)ivI6&qBpGh#woAc$X~=HiTmV>*E>6m!W)K9(5B?5#Z%S>&G<})EQ7On#_-6q zQ~!7FbY zFDsq?QH*t+*k51mc2Cn1%(~aJ>mq%}NztN>KKe0Vefic;&BT&9{al$h5TLI6Atj*s zd)5#EKvA9{5pa()L(UJMA0FDMd~igR?PmJD8q^G~*pH5@xa-U$GzuG-5}7S|932fqC| zT@TR0q+UATT(lBBnI=nN9meJX-O8u-TgbKZTd=)>JM~D#epGVqkPL% ztY9UB65!y883W(6x~z`*dyq~?xgbAAp?Bq#lkVzNy)f%&N(0F-!jDS}dX+~?Y)7x> zH6lGq?0zVq5JS_pPhypskiOvc!j%B4hAfgT0FlbZ;sr+_!gQvoDn+u>q2VHKc^8;YE0VO_oiUf=f zfPY~1oXP!yZ6EVXGi%7shJkFIfO|w^#-Jb->hlKy=%OfPC@w?rlSV+Pb?Z_I5wSCr z#{Tr7WVEh7au*2!iEGmsUGH%UP6`&SAdeG9hW*m}&rY2766J(H^KKyaBwV~-Fiv@f zvHs+;Y25wdzdAd?4^a#UYQ1TeqGh5ziY@H5>AlXoL$18lY#O_8ILN|aYSW+U-)0&C zi2Z;mT;HBUoPLdC${AO&m?H6Ar=CZMm-NdY6jE&&*G_@7J!HfTe_39VL=PsDS6W#A zzD4E*gP*LiDc<7vS+;`Az9Z6hY}XkU3IC)WfCrrgkbTP<#d2%dQIB^>EY`T7b5Hh_=l;IlI@TuyR z(a7O3EvFB`lXs^bi2AVz(`30R#%VKpv!-a$h37D&HKE4m5KBOL`G^T}A=)ClJOWOT zNl+4*%7K|^r+?AOjP1bDVp!Kj?u1V;1*a=F^N$bfSsAx~4nz|YiudW24cq7qAEIm% zg+`^Oj=Y#%5fV+k%sYB8)1yhOO6o59Rl$s8n(qb4L!Nx$vSUuh_4TKqI>6L*)bPjt zz3NifMCR@qIu-&29kh{0Wi>#hPLSvu=xhz2ygPp?5!zQr-R}=m6(@>fNxNRvNaKdP z_su7Vuc~)Hr3x7nBg&l{Pc!`1)OY`) z<`-*OYAAqm7?H3M+><<~K2T^2m?=$o!IJEo1UH@0zCEYsiA<4WdPY*gpSMB0n&CJuMrXeY*?%M#-n5zUJD&5*UwZcvR7AQkB!$Y)Rz_1Mvf$|hn}LB4D32#q zkf}E#wf3^lstT%w{G^_DYRLQ~?RjC_VtJ6cDMFy_Q9yK?&xn#2+8$l#D8T!Y5alWy zF?L@T<8WzzJ8w}X6wim_XuO3|$kY0~#?@WXW_PbdQ-;{>P9(6Rt#6zJxtI3Sq)otpR@f%wmOFM%D9c_O+ycj8X84!nhH+ zs0Y+GJ4##XcZiip%`@z-9=0=3aeH@f#*@{D8i`L=i&HFNg%nRG-e-lWmF*&%z zV#SppIxIm)NveB$VDCUR072fK&Es!PehNSo7Ms8Qm(P;%M$(utdfcczH8iTt?4gA> zqcut+c=ad47;h%-{vUKhuuo=&ZYp{(~be=9VfQu#c= zzgk?Bw9h!-HK2k0F*qtHUxmX+-YQyap|#$|p0W2Xm*7X1pSAp8sU(uU#cUt*-(Ihp z2Ce%kEu=xeFe804TvZZF6g-=F9?cY6aUAP^z86w%U&KQh?T@EK#5cTL#l zsuRoeoD1_WildIdTLjj?!++t#rq$}A5*7BY@v*bUZ<8zLk|*USv2=6znx63`=OF8of%UmHqy*V|hm_G9$6;E4Xu;#)7xwEIhCZD*LLwJQ$yCj z*Vp%PSSQ=|4xl98;N|=8L=*p|h}*lcI%a`tok}n!<}q`0;VA9o(`;1`U{hGKnZfin zmaXT*2QmnCp6UKw^2gHPv!aT*vtccR6Qd&^nB9sx%i3w&=4$p$Fycq@3mN3l>DRub zLffdbmdEneH)9G!0Y?5?azfMag{WCv!q}6YUs4iWiG z_1yD4-puO3Osmy%FOD%oPXl?Wrp(OK4ZSUF1>~QxDfG-1vQAp5UA6@&gENlRc3<5% zSj#nV4`E(Ob+!(>+C{@6_FR6^@uWtSfq^nCHiXh9p4f3x*>NdHLPN86tpa0?+rOKJ z>txPN7<@wT$M?KUa+LcpX~OA?DYOFGOFL;H#K>wkz99^~vfwMY_2k86?xOui%s{ei zI4J7bPbbLO>H4|aLq)})=>1l6iF$_B6F2eSwT?Mk>^||$D#J8Kw_xjpNomM1rc~9> zK+3gf%4I7}WUR&mZr!Bgnpz~WUo*X+R6W8%_xF=qJQ%H{gwncI`=SB&8N%KA{@~zE zKAr6KW2*}cL2hOOQ?v20F}eg?)b}$rZjeBg(#mhGg8Nmf?3n@q{WAZlrmCWsU#262 zpN?x9mWe)B1|`4Fo|GSp{8rB=ApdaKzhuB7bPeJ<#N_NxzHz6=bvyfJqcr71C)KCN zenTexliIyQIYm>9(M6nQ(0&&2FAFO%?%Synz$~)_@LMa?0LNLdvRq;_l+>r)< zFz36gK2oheQZst+mJp0KY8uA{tstFvVXftl*K)>?NEUd?H2vkIkU3%`Gd5AaMywz6Az5nb0q=66jJIUw`IWV>Ver1-eY~q3gs+ zR@kS)6jAt>Uf|CDlLq!N!woNOS*2gpaSl`4NlPg^rAxO7m;51EYPwKR2VICQ`zmZ9 z&dfzg+uf9t*HRZh8f&VOXAt9KIL zXGZJ>M`V(HzJaJY8cbbc8x+Lpnpa7HHf?nj{1V#sbF($cU;?85r;m|Pa{c(kZq2`x z(Cx)Lx3dcLkveWLF-cnR%Z`ow_diTU$L4a{p9r|>_3N%-3YK3lQZ;!c_lE6(ir|ACOa!K~KWY@h!(&oG9(Kr#OCn=Cst%#HER*|8=tqe+&Jn$J z^~wH`^P`&3zia%)^?U7SfNthTUVI!bCl8kkeRISg8m^UTfq6;f zx&`xcnmWLZE`^Mdb`XYD9(zH9w}ZtNm5ew~{oJw}oN2fW=f8=S-&rn+(O(dm%oAn* zLflB>?u__YTyNTTtBf%sP!*Z*E(ua}_x3lUs4C1gg1S_)49TF-Pc5tQs6X{nNM9Lk zWC^m@1ZSX7Bt}VxL*67{ZR z+okBItWulPGOFcr>0`(8CyB;=GDgf*u%fNOl;1`Qc|visU*I>d=|aDgIwBGvXKzZSycU$A zAnJxELS@(xPqrC)|LuWI4=*l1b7JOGcVckd7e%_T;@$BnaMnr|0&(z==%aXDmJPZh z15}$#Pc*JgOxOubP`|kj#fWt!8JS8%@c;tD7P#@uiJ&-&FmBOj-JtT_l)+!OG&|6q zMaSH0ge}RdPAYg?;9j6n1*Ddh4F{hR?#S(fgpS!L^K7^ty`N?W~ z2_F(S$gE5QbYnuF(WKa6kD8N-3=vbMp!2}$7GkgxF!x4WxEXl18MV{j)6F&(jw$!S z&HDp^TCWfp_j&WvgwdaVoZ8J#jdB3A>Iq|34Y*&qy1uwHGDMkquHSUUKn+|-yk7a8 zkHe_eWxetXJmo(`*cL709_&KdQ!{kzN$k_LvgwRA#*FYTJ)QX{yUs9(+}`cCpzYZL z-Gjvx$Y|KN;0y z8<%usfEfxikvr}v)=RZJNX@tjgdzkzWTHk+E=ix6(yLd|;T=cTo@TdYzh(dxt_|5b zuA9YFY3{D|YdG3QUvsztfjj}r!kv!F8^uct4eW35Gy%6kGHQ|ke%)TS@~{sBQLyL~ zobq|rXc3JfdqQ8cEH|0N&n&dJE4+H}9~p<*aT<#j{$M8kWMKEW;$=c;pw==UA1~t4 zcmUJMmajF(wXoD?BHiOkvulCuJU~VRI1tIu!{=6P$P?Q3lK7i^^PC^HRdHC$`kNrF z3YNn0_ez`Tm0hYZ*!34d5}W@J#L(|C|KC#F3gquz;{_#59plgmJ{jE^XzNGx@a)&> zYE~FrjPP5-Kalaeg5rHdKzE+RVn=H6v5ybe_pwD}|g0(L5pXl{>-K@PSiwnY>bYCEgY0kmr&763R( z2Q^z{biXP!$`!FO2Z-JZN;; zLp8P0yD?(aQe*)~CPkmgH;5SU3@W|(_}jJ4kz-o|3vZdHBTOA|z>9115;7~_+{@~# zuZ-BhN@LkWh6wnBIUgQY{UYoNJhIH&2e9}?2JSO&I6pUS1H!+)2QjgzwK_Qme#PMx zfXFDH3RI)3G?X+0F@@${+yhJLtFYMFjug)2@Ng~qpC=OfS3;tVsR;~Irbbq8HnEL_ z2Tlh)>izwu*aDAVUvfN|Z$Z-7WCh(RUp}>k&xv0aEHOP-uuPK5S=Zfo$cq1((_EMb zGh|I6i;!Lr;)xX9Y4#50a>dKUf5JBMq$ccB@Sk7=Gg6jIR zt)ybnVODze#l6sfBtMkO_7gRWMmQo?*CEkLvIa5Xf z)g(a-db(SS$_f4sga$G#KFSYH6qT;rXWdlvhNOT9KFZxsv3$Yk=4?9D4h2%)a@(=L z*d+Bwd`F(c44$H@9t68kW1wR1vWQxn8U@NBMXH86`e9G%)C-1F;q%?`Li(jJa;3zo3x#b-^`M#+8^JLvH=w zD6>Fm&HcA%>X2l`-d^FGIOy>R+;uw(#YV+lIR# zd=j5sxW4KK@yK>F{dxvyf3_~ndsw@wT{*6$y5zB&#NcOK7pZ&VNjzf-_#X= za<5MD(E{kG6)%q{rhUBMZXZ{ayZ+VgExl{he|L9zJ-!huspC&#BvjeEGfwYm4(N96 zF|tr|$aKgy8Ht`PO*uWvEv`RG%AynBZ?%}|8!1UDfD{T-raE&nHXX#aLWf&~IL~9G z3kyunV^rDSys`&c0*_j~6poRyx9998I~%t9ysCvd98R9MvQ5_Y7Dzt-_7AjtZ}}p! zvr=hfL?lDaoOgL&Fm2Wr*w~y-=>+OPTU%TRn>b<9Q!JP} z>$+n|2lWKxyVh>Q3(mQA{ezoQ#p|RO?uxVVl-c@y$4S)n$+i43`A8Gv_0SBhKsxzv z9;*w7P|FoG#xJydVeT@kDTgxOpD*B*|b-_x_=p+Zp8X>p zmvK-XE?S6%Q7PwaKT0q-`4u)*=wYe0DS1lurgV%-Z`*1ocdY?keP!!Nae}9PoIBF* z06KYoFKKOxnYO>O1d<<{|G=H(DNa27yABBz>Q?J0@prLevsUze``Vz|AG|iYkkLnY z^zy#>+~3H@`T+ZV;F`uV;9j>V;5%j0{e(|jD&zOafGRr>mLO@>k&vYP5E3`})OaQL ztLGy)XHV(TIPh(@k7$S>wHO!VgQ0z2G;35Xx%`mhR3^)4cMkPs*1w1?@rFdgsm01d*? zl{UYkutnhHXe^7v!2ij|5Wp?avk7#!MjPZJXNg{|xt^x8_0CAh^WqR#M8-04rx2~J@D8!SF|C{?QDA_|!<=_O?%@$+3vxC?Sh0ZoPlO$G)k$ZVKC zULKCBt;lWwy9uqXzWfs{%^+QZHd%tUNRZdSFI80L zQ_PygoMw#RHOv6LHJXLbF2{e$Wy7sT^8}w-bf<_ZsZW} zCfE{jm6A44a>cgAP0(FZSKfUBnsj0&{Gbd$jahXM5htpca0U6Es@8v zfv+*g?D$ElMV{>GAE$l08`Uw{I)l`Y=5Xx1X!MTRci`$~{Mk}+HfR=11(J8yJhTme zdI?0D-{v`y;fvxPmu?)HNRMe8jO8{kDe+2~KT*+H0`0MQEb0Nz=$rv*?xuLhbC%Vp zjs*{u{H3d@CAd#*i7sa;0%I!BL{<0h4Irua9ZNh>io_;2(i}^P><^cXCBQ9mG9gNL zmY<+f9^Zn7LEzv;2DyK+^BhsE2WKHS_Tgc}X>p4k&-`iFwU7>#|5-AhP`7s((MXiI znChK_M8(d87K_9tE)vn6B{gjt0jsWiihz&Xbp;lEHog2(v?a0z9Z5%nPmmgO9TtSb zTSKe`*xFn``gK1eFiIsaSG!RS>O!PPR+~?tvb{Lr0nl;cf>MH}e_PsFF!pBYqax3w z541m6_Cd?h2+u1DbM{6NtkZ<|CJU&@6FNi+F9x{`h)%OtOWY`u@d?jnd*~nOR8?I+ z*>M2SV`3Td7VNfAs|Bu8>b19~1GDfXTECSsb9;X$i6DkGJeht$9vSBf1M=&O-ECT@ zZY{;yp%c~@LGNdC8eFiw3{py9s@kuD#k61Sb{W!@;v<0I4|(nLsLYS{Yv|HTg5P{c z_m@1^Kr>?PLbmpkU7dM%4b~*wxnOno1d$J`(1{~M+pVj1buaWIGro1(8uJTMBRfMB zB4iKrM8H`o#+^A>;6MijLKhm-K7NY-q% zrEWk8_hu;ULXA_Fl%1$A`t5rvBFCyr=QYlbV57_u{j(*iGu4 z)S5zgIOpaS+_fVbbTsi6QQsz`PQ)RyS*L;DrB6- za#r=U7P%x80;RXA0g3KsVywXd)32gP4RxGAO}tEvST-?D{vO+pZ^RwkDLlY9)kChP zL7z>27dN?!fP)phpz*=LywC@VOx2&!Mn02mq53P?qP|KE!r}BqlFVb7@5`x5Auw}$ zl9T_383a93hCNe)w?0!sob`fp|1WZohHwDE0(xJcSrH(ESLHc#c3w6^hUE|t@Vh1f zEBYHLZ=r2YlcaCXy)bRT(wIM%Jl)lAfPPY$DoqyhGr}6Z)xA1Cn z0U+7D_+|4G>q}R2EM-7ipyFkH1b78p29mHEx;ql`8pR;Uq+ps!S<}|MJ_LCOmo5u} zEfs*vA$lo;BFF*$VmF3V?+{4)&v+DEi3$iqCz6IJTz(TlJyD5Nqyx!kSXSWAtB#SQ~Z$*Y4j^ReJL>!dF1DSz6Hc``rNMVLUD9#`SZq+V62@*9A%BlhKe zb#fQeS0}e@+UD2gFEkoa*S5c7L8rt{&R#0!dkWuG8|bWO<5gLIGngyII+G`WoWS22oO#>SIq#>MW9EkeRM`>zi^9mz^px;05)HUQzjN znr_aBZ@Y3(Cg~m2e9>x$Y;GZN(WYT3Y}qEa#FD+hpB+P{W2jL3X*o}eKvT`MZTqjM zNPuAxu2-rYzfZ1fvK(adUjpc(d!?HMKWIqQ{eCR+XDkHy1UXMgT>;cU-QF=1 z$LIZga7>G(3t#^X)UzLpAl@EN7ymW2AW5Pp*yE)lAl>_d3>PnEqm|b0T637E&`05} z^ks{MuQv2rLV42V!Mh*w>3?j7h|PRS3`a)7HsZP;vppsrq*AP?Ra+ z>nV^m%;okv>g(X=_j5e}2qtl8r3bS9Pdbw;$C1!imSr)vM+c=aIg<_=h~6>$?S6mE z#s3NQ^C4tPWY$!Wz^@Hnm7~aoBnPhrygB4A{hBSfbwKkhVcvUqEB^3_AHr<~@^ms7 zPGH7eu+=-(w%vLuXb2O>Pj0+0BrcixpOaEM)nh{aB7_iozV;6IEd#W9yts&=qpoP2 zG@Q?SI)J|SFL&X+WXY^I$1ls*I^BE>A%^;bD5+u?GX;`pDDT?x&D_Qb%iR`r)~~}x1imG6d&ds5um(M7e~z!M1WUPKmlldR&Z$L|9Ne> ztFX5;z8eVe{#uX9HF%a>GS$xcR9BJpWa3ai4B2~QTq!$>E?N{m#Qb3I_$CdBudKK< z3F$an=EG_i4e3N?g6+s*lC}s~3{_X?9ho(eXl;8L1Ac?=A6(Lwf4j#7-ab-N5lP#MPg0M^qjy1nhD&+*JfIjHRDJ}2NSTcy1wYKV_+Pp1-@<=Sy>0CN ztpIL}%qpNE1iJo&YM1K#4wv;4p-?85tOyrnP?agaw_FNB@yb0vwOmDFfT}w);WF%J z9Z9?;i`4tA^Y1fC|>$4F=qQCLc1ATel+0K=a|K)!s{p%&DL_)<$>c|uI1OTwJ@ z3)}S|yGBP;#e{pUT*pT#fg=y)r=6|tAKBNEravw2tu^-uyz+QXl%(8h%SZ0fS*^X1 ztL@E~z|M851UUGw$!XF0gwUuRxxbEeHu+;85yX&K7>T|WHE}e*h0t*9D1m9lb`gbyAK3i2l&QA)wuj}6U zMjy*&LwuG#M}4OYd9&Lv%YER+8z{9Rua&jDle(~#*j#i6i~t9tD-#~*_gf@zi@ z=LplZGUxI>m&;F~=Qh6D7|ao}uh!GPybNwU)GzR2bau<6XWCB@0sJRkHK1U)6{$-M zWxXh|1sN4_g4819Z#)TaxVNsH-0Yb@Z4xg8544ZNhd*Z{p{yad3sltp&i4P^7Rsfl z_N_4QM#rYlN<@=Zk~+jhE2D_if9YZX?a+lzFbA>E3(5asdnw8BFWQR>_l=B%xBoj6 zt-Ppu7`)K_>p5+E_^>JmWjZ&KA!;70lo{TdC`*-B8Jm=>QR%R0dp5#bV5Ys!jZP2N zPK=({PEDs@o6^O5s~NcpJTqh?76v(l9<Jl z79;oNSg+%ha8xoY2HOcu36>~$s%E;MwzzWeHTj|9m4PGB4&Fc|MT^siZ%P~?ei`aW z6I*T&;1pAkWEZZTVDB3J!Qn8md>5H#H~?&3&0kGcGZp9(IzFgbD&X0Hgw`Mit0E%N z31!`9MtU(P$`x(VsvCvDq+Q zz->zx_?ZKw4(3!OXuJ^tdE*M3p)=+m27t*a9pFQa6L1e24O3Rlh5~ksipb0c)#;h= zX;=)!&iK7W9q@W7ee&NofY~HbXU*!e3T^ACN%3Q2Dv}9DeVE;IecVQQI^w3|%$6`( z(ge`Fv}?(wJS23oysK~rMOVtTOVFLTBerA&)R!22G?H9DjmL&TzGU#wEx@3AW>a&m z$P4(`E6subPLR<2lmZ{a+7&HY@~OLtb})(2h)_n5T9TZh*|~H(ow-1{8F;|Fn>IJQ zee{856c>uq4~*}O9~Hm%$xfG`3KnPWq_4-ekw&HIbpRc(VtnuqB~mJAv3;sXjxs#2 zd8~|9SqDg7G~P*|+*0m3yHakTenm4xQouZX%@$7u8 z@|im*c7DWa+xY2G!RiY)LC6!k{U?^Va=y2AQkyrf;#5u-oPAwt%)s=*rm8 zuHP?*&S|RYEIibWc4K~m8m(qxvTEZ$!1K~TvJ;|DGFlEkIK5<(5~Rg*1r(eIQ(AIeODdvhh>N&DI714rk$weJEKwH;2E(SzmZytz`n$yK8Z zV8K$At}5){3;0WUMT-Qd?A}*uj<|QvTT%-h`S&Q&L;^Voy!60^v4)d+RelliQi?|= zjpMGhQ6^&3Yc(LeQr^AWT>R^AmPu;JE~pxzK!EB8^^@5#QFltqLAUngh1c)3UlzZN zp+zOEQYdiQt!v~U%8z^N*{)9nNhK`&J1>fI-^55gLe7PWK1}6m%#c$pAo(gITFz-2 zt2R@I{HY+&eNmh~+Qckrud~TqUCdnXH<3TSw%Y)odJZwsOQj_Q5-|P#lg|c3ofCm;tki||zSsPsPZTb8D@hcEd0TlB_45Z>khhxBkaF@~;fv-c zC?YQ5MRTTfPd+Q^elqIbnS$X~CVEIyUYz+%hMr{zdUHN&-hZzkOZ`zu}CLTnY1s z+!V}vF%*6%DlAZIVQ@(|()uYc-INvTAo;?({VGPb#Kq{MMjApF(zTD5qs=W;AyMwI z&$ZV{R@dB`it&rBREv8Rkazc=Wb(m5aD4(f8NKQv)5x0m$k=e@sx2~#m6Z5YJ$SDB z+dsLX$dyONiw3O_9;Aw(uhW~u@%7XN@ew`SJ=>+9x?lt>-yK)YMu>0RIwuJlpynzZ zo`yG7Usw~sZ>emBNkIWK8c4hpg~pp`^O$!>(kmtXULfGjEXD`Kf8TJJt^Woh*y`>` zcrK)#SxP)-X#@ythOPhirP1T%G7DeO>TK;QawNYjUu!SdeDrYc#A%Bw#8uM1G=6_? zV}Ue`%v8z2YD0GaQ+YEshb_0lg@#D)Gw8ISa!Nih;WE!ACF^qhYS0tE#K*oUScZFt zdMB*9YbJc4ue&|oXPyk0)bOuniw>iu9qc5Ri~O2sx;rJ3ibg++pSJWtNhmzOnx$hj z9-Olk6JNBh1{dbIB7Tf_qrZZo125%+rf+}M*JdxmAj=#Q#@j<0ROZ5SRFSqMYQkwmE?wgMo1K6fCdU;709Yo- zu`u?`j$&R=BBY5xJ5)@ujyc%0t~_#TuA@0+rPuip=AD<32bk&F-V#UZ{(;IZ=0X6t z=9(sA_d8g;BQsbJJ)N6}Zhn-O?BdAUGKmc&k;a3Llk>YS z^E^pmU~f6t>lrH8&WX}!5V1y@B^V{pwH|Gw1T%Q=>tWOFW1H+M^1)WqMWwS~+ClXh z>~zx!xV;Ff-YA-`hK?_wX6IU zf0girI3vOIRY+r7iJ&t7S^mag^w3}>=8!op0wezA$7PGN9b?-|y3`Z4(L<8CwXA() z#Nb}PQ`zl3%b)Y8B5N*FyQWdA5_sf{tG?2%-kA=%o<7%moBF-fyjHx;qN%s8_l69) zr9z~(Z{1SA262WUf@>JyLQdi&wf{WLd9+ zdx0QR*>0TX4{uD&YMJD3g7je3`2&XFk%s7`Dbuw>?=2+~nQuk%pCO3;Z_;}44Dk)F z%EiUb1_1DIu!(X>u(7cLM7c%8Iezepa&wD`iHov}3X%c;za@GItny~|7Os|L?7X~W z|J&vzo*!okGQtYjb@?LDSp+UtNt6@Vyk#F&!vSl`aS@SqxfD6fIH>LPX~m^pY}Uvofp# znkxLTT@49A9{K=UAWpa_W0MH~gK(ql!#s|wFs-WyQil<=w-NGN9;vz!7>p5dC>hEi z8$=-+%syK_MrSK4?XN46suDFWTdwLezt#+u>aO~XUok-KsdB+?&8bREMNQ^17e7X$ z=CbMgw53Eo2MzT}M~G5Km}3^1=XihHUx{L}9mfxT=RKeP6GHal43%jy?+_%2Tx^`& M2-MUPN|Ffw10&cBv;Y7A delta 14475 zcmai)Lv)}G(4}J=opfy5w$ZWeq+>g8Y}>ZeaniAE+qOEMZ_fNP+u7a4tzDfux9X{c zfAB3Y@bP5e%*;e-`BY$(K#G=@^T9Bx-%Fk06}oD3FcWmF<65oLvZQC@dYetjeljPr zrW09zR1m1p*m&K{$^t7)Q-G;!kBqvYU&7zpkBKX+6~R$c@_Wrl z)jGQy^+FWK)N-}d(uUNbbGObn{iLi&3r~rVr(i8KtTD|FOmAk`PYw+aycus81PzJoww-vh zDu+5hE4OLoyM##yKtYfTZreuw{&4ke{xMhA3aF9D*UhbXFqA{!JE{4{PCAmPHB3G% zX-8^8NcdR1mLM)w><^B(Ilkw%f`vI*U0FR5o7i|6Ive-OJ~S^WxRnmL;9Na(lhh8DAFtdCP0CQ7YwJWySP*b|c1fNL(lf;A zNw~geJ75Y1Vzj@jy>;yJ+8UdY$%u5^y)T1?Q%CF#^-p7$8z2^C6srAQ9;!91YVwo0 z(ceoit?|mxpr)}LAe-;fODv>x(Vi8-&}G(3y*VlDPc{YWkj zxDWI`ua~_7l#?Lf_Pc~AT_Fo*UFp?So3ZLj46WONG~pF{TZR1oO0k(<`|#yAPTn2t zXSInMdC6cvD!E(mQQvtJegye)x*ksNY84RVWtgB_1W{k3{V`FDZ?SsX}n1l5fFgfF6$&Z6xFq>@#XLz^LGf; zIeac)-!6`0OF(VzmMywvk9hOz-k1DV6Si^lJBL%k`HMn%Arn2bn}RS=lE*U_o@K&B zSU0_n{0>R3Ey?*jIRly^=9uB^;b?=3)G_(1$F3-e)%urR(S2iO;xKo-dBd$};waWM zDL)7pLu%I`M-3ugt0qGwpMrTQ{vz}|_go$Tp(J$NY`0I%tY@wy{V(+BptdzNtOe6X zZf%=I{q?0~jTgO{sY}#844b2lCD8n(8|O1rbm`KXH+nP3ienF?r>k+0(yr0(mklBY zZPGB!*xzAR(<%`0#cbh(3s#zDsZh^88nGuoV+j;5g-7!2GAK({mA{XmVy9;w4-(Mq% zTp{V7(Z~=ed3k9g?F{|dtrko&ma_%wBEvjS!UNv9rAgv)KPH`iqw?A5==TK- zYqHRoKZPvY<^B#*lhz*}(usi7h)R_d0)#RW2$2!+NN$YnxftnCPr8>cUyuL^%E0g# z>|Pgn(2Ep%u+(&@rY#-6P#bJmsGp#d*fQfl&u*Bm*h_ZaTtzgi;bmZSD`R{u>sey$ zP$7zPDjwm0+%F?VV@$G3bNh~gZgUWfMB@S@I!gcH@UU(xC}+?ZBZ$Nu(JH-}EsgSn zi*dd1amoXsr$({BlemgGKpfIHkKn? z6X)KLFb6n0KY4@_s0W#3c>ahS9XB*g9^>&;@ONRx5Ju7@depj9Y2Ns)63gHOp>{C+ zD$$e;AT|-;*txKcD*4Tj(z8&z`P3e8v29hpbh^>|$P*WNIIa91W&}`Z5FPRjBfo@M zb2cDQ)vm5XsoX(O{T9)wEyN1DX!#u%q{d`#=uB5gb+TTYgTbAoZrNvi%fB9MvLTs% zJwWEzw~O6Qc^+oimG33P5S+*|LI^gpQ%r+uJg_5qB*JeN3|{KW5rcYXJQhA{X&JmW z1#uSy)gxWCT>47&Y5^lHaIc|o%@L&Y<)_0W+(+`AJY)%GzQ_aE?uhPM_vcivV~Q@Q zCn8`w*$bL7+vN_N=!w73qc`wONTTVDNe)=mEK-?BDB69GwFiSilw)OoY)6-ahI~_Dk($Kg<*?{quabFA_90O84)OnG$h)ae6E}t zzbZAhixEY}aMtQYNn3YHuc2^}?h=_CU~uQ}V1F{JKF>a`-_yi$F`oQw+8ySx!DR$% zltbrxtrU{lgp*N9dR1KC`c~(^O{YSmS;tV$F`s1cVgLy!rkgL8B(cI5es|k#iPG9a z5EN>l<-f|q)d5yi;(v*D`IcW68Mu#3WIBnSUcm{Z)fPpz`PVtUWGn?mYb^)O=DB#? z`a(=HXqU#)BFrtcfAjY?9S3|i1Rg`bU7Ofu_38Y1fUxVl;;=dXRIsK#Ct&2pqNYw~ zW)0IW2Z^-Q>hK}BYCDmI?vCYB!^TAhn|6Ss`En2wmImVT`o~^bwGVzFo)BWzDxxbL znJ=)v|EcNEB@~6Nn`Q9kf!7wA?DrV^00l23CmmsM7~HbtH)ekH;VCm`5Q9%&TxAL7 z&XrBZ;bt7ZLJR>F;0?riHlcD?(#Du))ZA$JIBJ;Ps(mRUjufb}mFkGOCo$ndo6hO~ zqjN0;5(apgUX5}O4W5>|guzu7&D$bxdRB+xx8m(fw}DjJYuN=YjuY-N|H3w4LIdSE zRBbJ8AdTe#q|{o5kg#gW5fHtUY|}%uW{t0Hcv*r*TCT_Giwurs%NFlZQYA#!{y~Pn z2smCiYBc-|!vdG=%wUw7PR%Qq=BSWnu826Wp#h2vu29hz&wQ!6;`Qas_+6s&s#(Rk zV(DCXi7)OEv7RC;!e#jIIeeR6gejmjQ7qZ?ZZy>Bmk`DevTrxJ8C?dy8{pyd%UnM* zK8quSl~YJz)qEnts@Q^T)wlZI@AC47^-f%U6{M8J^2WVsY_rE#y)6?;NAE=)BzTYx z9so1VjDcZ69^{k3$pWK)h5j2{h7)>1j>-eYC4EqO8=m@^oz3ya;HQ4_*^@@0(UaUk zS_*R%Qv*x)4$bgV)a8G&E8;n}I!*hwwqbg~(&5ePk5$IRD1}RIaW;~Ic-w00DsUWq z1SjO>oF?U+P6F<7hinoLA}e5mpM7{|JOB-1EyX%9I`o;>naZu=BNbPwEUo$l=Mw`j zo$Y|Mx-M=#{2G?vSgrCig_dM-;`e|b9d&X#B7I3guo(IVhdh{9$}>-K(BI4j{O1iW zy`F{&K_4ZLD#zkn$E5mx>cJ1qHRC`KOp0sf`sh z)N6`;SL3UbD6{b<&?>xZt}SDT-OlhUz^R!bxXQBT)f<~oWC!kK2bNfl&A|6{JcRO~ z?fd@i?e30QKvi$kZDqh?RlTLB1wc%G!iir@J$Tw$5VCH7Rxeu_{dtvV?QSaL7xZbd)UhrkdnNK`XUFf(fOIib9nQ?o)Y-+!%*YnbjL^>1-uO2h zD?1Sr(f<~Fd~l3%W_A`XmPD-VTP4OK4 zp-gdF<8~mA4b_TFmi|3*U_h83aBNLx>xook-e_OE@vzANs(ObU4U*y~K=6qUnT~fS?%*+kF_khEnsKuVvd(Esex%g@iqfqs$JkNxB z7t$;5_}ymJ8Dezd1ib9LyHknf)`T!|sBw?oD!S=;(;>;jR`r4YIegT7Jj#(%1ffl| zKW||?o0`49A=Q`+8D)THw*JflW=p9(uJ}&!h7_CS372V93YKL<3~D@j-V07c30jvV z=G?ECZTQu%EDJOTLvRif1t(%<72=4L)Y2YR4PWR2g2oaR{xLp#A3`M=7Gy$9mQRs z1P1+aUX#e`ohVYO6`QnqV>DQ6`AlEwQkxXkLi--@MrXscP=T&>sk|8W1L^o~G>MJx zuA5Fllf?n8Uiu_9n0?O>816b1cs_&YXZ6gKZf+2ejUiP6+9%ZJbB!QozrGsySGZ#C z#%B3d%<94a@w;9X(8fM;;6bY{o6RAsE^&}AH6!!&o2)r6HQ*6!~ksnb9*? zMpk{usy7a6j&o5i-?<19!EbDruVI2$Rp#ThK}MEqE~|ye3QI5pK&X=bb_Ib`O=Ri| zx{$w+m}odu@)00cCzaxlTM8Lz5qjIrF{l<7G!sT$lXk_SDOce~8QhC>1K6n66nU6_ zCc=o-cVs(cF=+XzpQI`Mo0{No%F3wPUb#rQ-LLnuG))B}g}X@PAf4bPKZOX-K3ytr zZ)crex_dqI&I)#cj&qMn&p)N6&G1L%ZUjpVG&#E94X!NBxieSTtwj>>Ve_!;)X~=< zr30Ww5lB+ZIHFUzz`H+z;Yt#i=vC|>-0iF)W{w_Lt_>eSIWTOo&6wsdo@*Vf<3oic zzVtXy0uC35oyT)^0tX|CSqx2YH&Dh4g;@T*^BSx+6Kn%0qGHnIVbGigEBJvJ=5wfl zOFBl6a${AOzq_H?WFHmqL;LhDAs_ruEP%40PKs+Va$l=-N$v2lQp1__4r4I<6P%l0 zin#s`7lu-?<`O1P_w~3{HC0#ex?$kA{pItm0HQV#6XD(a@9b1)6DKoSo^{CkJ+E_9 zENrLX1>6PLk*pO$>YKjWDs-8|Mbd-EWlYK?0H^4yTmMag4Rn=q&Q_Sk>U0BNc^my5s=h4>f zi`~!qMax5wm5LnVyL2<{+mq>4NGc$yc_~biOqdB^Y*fwrqfd}0l?mwysnjg77C9?% zHq0)&lHnR9iP%oqNA0TQ;C_Pr(BUikjWt3}1HxE()DO)?CRae8{xK$KjELr^7wrR##| z9N+`ouk#B&Fs|l{l#SCJ{*9`pLZ-wLc9t)fOD)kqL}d#*U&w(}8=40FcG~fW@6LG*G5kkBVnhcAe{2n3tD)P;_HOs}Ee< zP+%Zj4n&&MVWYUYS+G$ltUchNfcd~bWRVI)u3tcKf=o%BN$V4ootMMt*Om=8PWd&c zSAiyY)VYr?^s%kg*P}QPV`#r2N@O+684sB@{bR$S7l{7KBM2%(EJA*q5Tkjt79mw0 zVj}!0XDMJCl5NPA*GY9NH(aiY&CbF%0CI-S;NlQoKf5pF;uh(a4$@8?pMj^-lu8PK zQ&JgS#uAu=+6kbMk5I%biLu^0!sV_R zzh;Ln!v%{byzwvTPf%1c_~z`b+8S4;ComMeHT~FOs6<9*;`7 zdrrSZrU!;yh&O%KLy_xIOw~O29-ZPZl%vY zv^ZY#5t1cz{S;`jAgYo&sYnW>A5cSVn+#q(Iqzk^xwDA^n7;4S{d=e7{yHhtY=A9Q1w?P2t+zi&s-;!a*;f4F&RS! zl^QE+U_x`y_T#X3T3@^zU7-o2JXX+!tI1?ZY9C3?!O6wVAO$91-7JoeE5R?efv zvy|v8-O~j9@A1McT}20l2VVT1NNI74B4h_-GPP>UHEu)P+P@(=*>X?VIopDwYe%&x-w1#r6;c}DjYSW zMl|t)Hq4}fg$?pw(_1IzbWMCjo0Mru^?%p>MnctA2i_dMo?{F~GoKzN9@NJ>=3w=T z22me1DPz#FAI?eSJI~<9#9#xhsWDKvSL+WI$G?C6F#(5w_dt|jKV8AXUoOSJbZtyw zc^UZp?M5iWjL5JF*^8&bp%gkibC3F%{@&f5R)eCx$FIKQQ#9Dz}@8~Zgd!(4>DsA z{e9g+PkW%8K;OyqdvwO8IjI6lu;X-`>~RNRo-dprcdEW!O@W>?NG2AJ#|nTYFtP zm4h9|A~@-c$In8uRo1Ux(oXj}WaL{w?V?4$w>b^gLf`>O6NJ$Cp%ZE+uUKzi(wNudv? z2QT1$j)d1IKLvZ_E{K29nq-DB7JUw$_QyEFfShsbrsVU-d5f$P)tIeaqq>$5sruA{~5?Fd51)Sglgjtm68*^qD^1-{;@&*7#9Ls5q1m zZ0ubBvnh7{hjTV0(Eh_Yy&HklsKdJ0MD_`FGRBm(DJ^a)cLG9#;bK1_nz*V@cgnmw z!XT(*Vyl$jNhbOEeSJRJQzSO1Vz#@aUcEg(_m6a_3rVP1GQ?L}6BKD0vr# zJ0kt!1Z$I{C;W74VkQr!$`-?KG%ols@2ojH`~ZX2*wWz~dyKoST0D>HPQ6KClM*qX zI7zpQ%I!|K^3E0;%lnZ6yVk``_bPU}O5LRDFYnKViB^&<_{?Io?Rbg)351{wj)=*k z3x27O8#M-#HU%qO&5Eot1F@*TI;_Oj{KxtCJAiLZe5tWvztHT70S0n%9+j>`&hSTQlZCIf*QXSb2U7y-m_Trxv zNv-Ecnw?8i%=ERug2!okh83?E%B-0a418S_th}V)-HVZD z&47(3Z-Gy}Pe}X_1+xYrZ%YGEV|;yvUk159B7WFRq3OOGm8T_!w^i;%2R#}y02xGO z!AsW@U1Hi)d*Zu{x~7v_GPJbejJstlv)W=1!#{J|1V*!uPp6CCC(vYRD~Ed)$xl)L#YB~-xzd-;X$!FQ^Tloi&RE9^TCVYfkdod_o+(f zC!PihDkHj2(L-vBt`q!z!&r*-n|O`d%bN% zO#CONlyW@valQrYqoK27prl{Ji0MB9$4j~Vb3P?c)Zzye=sKDib1oMKOg02udZFqu zF=M!A7#NX(IPkK~3pjUUG^h-DHVmO6g0k!rJ^oDQsPL3(ntk-EyRsJ<;ytH8Y1Gc&lQb)R|nN3>X zG|mIgP!-byPV*Ww3@-sbos7dtE{n^J437;{3PDbOM6D+4Em5NhFce-DHZA?;BbzaS zK*uB%GkzddVNN!2xulv`4Hxl^jqGI`JeEIrZVLYR6~_M?3fY1JT-LvVR5t$$QTreu zwf$4i`Q*JVdyksGxsoB0Oe2y2tlprekNDbAXYE0jCd}$7mPgoiG4w7W9UBVUGLEA6 z_t|dxN5?U4bh(R#i%8+>bx zX%@t;-E3k-RB;ZL@ZiIK!0kKi+;(67C47B=t--SD z%YOBu$%&!$kwPL)gPlqTF5%WqU#+!w(3U?KovBelu| zud5Uk@5+hjGn@@A*vHf~iAD^jcXylQIYG>SRWTw*q+%5CC`Tj=qnbV{ekq^IT)8h} z(DzUA1=^4eK&~C4j3F3dP!!{qovFXP_2aVC*k1?#9`2OP4A2D``>r?vdBw~yoS%q$ zVK(y@_X-V;(RqV{4mp;dGK?hB7HkVmTev9b3#QJNjSFSF9)*xvfvrz|a9z&7eF<_EQXJ9I0D z@8$oQ;b$8$2KXll!*_O=;d~RIGgG1n_g{4?UUnE+kF+Nev2=nx!!QWbo6w%NZ{{rG}se2;BTd17t=WxI*k=i*(lnvmhKkc@O)rtBY24T?R%} zF%x}%+@yBf1D)Cidfy1m+1yhoSQdI{$~cQb)n&x~ZupuoM8;hinSUsoJ{6xu52)Rq zkV^?kPuaCLpYCh^z>6S6s09fIxx+u7C{)w~QT!*Ew{k}`a=E5IHpPlzgZlPYAW;bc z-8PuqS3;^6UzNiOnR&jn^Xk;xiZ1d+KghS0yKTZzPWLCb*6xaly}aw4OnS zYD$Grj7huIByJ0-D)%?sU+8D&Pd>m9&UlnWA^3P6?{6F_X8kEY!Jf|(1Obftov<|k zW)?QQb!}A6`SjYaF4Tgi9xQScNcRO!bM{{37yEag&AfODjhdVDd(_5OJ(`+IB@Ca^ zbb4O#FYEi7yDt$Ep8Y}_(J(7P)h`d5QRPuQ<`qLwJg$?R7f>G#6~>Ptf6k|C8<^@eRK2 zL$kEnoi~HOOpM>`f@xtp@I+O zSYnD4l`qyw5Im02Y1me+eWm2xE@t7p7^bedzU}W~Jd{UOA-!O)KzS({ni~{=Pj4pN z?3WLf_i~4|O{FD${6D#rHB28|!Gaw>XEl5fEsE92kljq#@) zOD!!^X*gtUT||^pGn}5R^Km6~(jPs99d3sO8fH8Hlu!+pyrtj-*eq-92!G11(xHLq zOF)?5cMW~FoD-j0**DZLrK5WQUTW8K#hj;0uFbAkrt|H6&TJm_Es;kd+N(MvZ=T*K z{+&cDKl$FX5E*XOLU=+O=fyuldj#)Zu-toIN-`uQ;1p`geWxDM%oIunx<8idQ~eL) zCsK{$BT_K?23J3J&sYYVEhbXGK$m>&sd(xvITe707!`9$lL=(68$ z(8UeEf6l%|`}P=bNgabs%&lwpdWdt%*AD1lvPP1HjMR*=;aiH76WJ~Z`A*dqXjm2GJ4GRi=?oKBc^{q48~RCUoH zyPxWwo7^Bu`WjeX!zi7mX0<)0fbV~TD+E4_FwEB^*Bu+WYFCJmMQ$MjGSBh@DO8S3 zz)(p@lO#nEL#$c{rYX)>AGtoI^xXvmbM|o>7SA3Q6@E=mRrmXHOjL-+7N9vL`bVwRVVnrsW9CGB0ma- zo@_3W;4>H)`xj#3p6?$t3^4&Wf*c%j60nZjj&3<9s+H>t(+D0eFC67SnK)BWWazv4 z11S_$93S3W_AP*gMZO!$&}vmhTW%)-yCLx}CYn~wt{ln`jwLcj;&Q?}s=_IZU+6C~ zjnvD(Y#Hj#ll~u@R`-7QE*62?uWuJ}cozE`Z^hGt`p^2L$=l|}J(<3vUtjK<+W;zM z8kQNw_T$Vr{#{}2L(zOpEn?}6z|IzDenzPhpjr*U{p!a>?Q?Z3T{$qxY# zxRZZg6wUOB_;d@LMNyA#VHq9fkAZdct-B8wsNDVhOdqtF?ypP@lo_bT%G_|1yM>LF z$M>aI8@D?_5udLQ?mP`wfeLZfEaWODn(;e`t2V)`M~nRfGty^(zx4FSlBuTOSxQpl zhYttAl`%N-Mi8~gH7uN5!h|K&qH~*|NYAK=!^?(fkbJgUeWmI z5yx?rFk2((KL1jIa_LZRveYcXmz~wdUK4pR%gUGIDh~Iqqwi|KOH;sHh2j{et;lB5 zm#7?pn+g6)=@4e@-M&)rQ~5LFzn`@^LEB&iCJ#{hv#_XesTqDE`JZ{v;)-=n{y6>& z?kr!FQspT#^yFA2580AJ5Fhp6^x ziRWG-(TT@#pnrrq6WHY-4d2$t7C_@b|6+8=oDa5ohq!6EQn^C=0r<+c7nQ^%)2%xy zRFFL(T4XP}A09;sSo}N}CF}R^bA;t3A-ue@*f0KI*?>5S5i7VdFP(iest_spEN{Zp z^NbasucwpMNC%bT%jvL6OAZVC0if90`;T+P+=@xF#14z- z+-XEM_I3!#VbE27DPU4a6gUlOp@g7Cdf{R0C30=85;?0z0)2M+FQuPj^#nk5nHj5f zuBnDpS-Y^zkRxU@{Ww@uBork(IN8Jccc=T)o1Xp6Db>|sYtpCE=WB!dJQ-)TiK#xw z1Ro8QSbPf_z)<%6CoyL~v-J&TLNkw7VTP|BQk5n015rExicj)|P^qjHZGSZvmj_=G zL)ism_!TDNP}t({%##b74HPCw^_wQ5=3P|7Auk9jUPDU#G^PR(Z<94*WPG}$tRlXX zuYsVIIN{?+*}`uR^*@P@idV>Q15Ub#VyuFn_*EQH!0s*rcivNk2C#T=g)~kO*vVPC zuobR*ZljMVt4oUb2=#eZe$Bq*A+RqkUstd;RgGilH_>cie!^c2LGllp35@cp3ZkgkVwh)+9K7@d~NqT9FfF#h;CAUf8+h$jw_9CSRAS+qrNm z3VQeln3p8{X2$kmt@!3RSD5amrUY~tNuqZ;!Zj)Nvbpw~#RpxRe#{=!a_8tu%EqCT zAFqe1Lob?5Z+i%wVUTT!nALEdWg8Lb#cQ@|#cMj}1)4VNShsRa^H^EqF*?J>ZHB5l zBL2awXhM41KIUq3yS%_VQ>XCEkK9TL^O+U|s*jzeDbmE;q-oN|c~=BJoa3tQwbPlP zPG`a^6A;<>iDrT{UPfRKXDDtI-h#YXE?HIigwdW79QakJ_7RecN&kd<3H>0?4=bFx zcwKvq4$FCBAYKJ4IYvooQLZCo7^-Y}w0Qf3xJ)JK29iQ!%6xgo90|Jduvuj!6WX8=1 z+DGP`O?K-GGj4TO=FbeBv$V{Zc@H1E@@Le37>A)d%Rg*jI#dqhl?koY##6P&B>Z!c z#50s&WK+V3%GH~?{5>|Sk@gsKd{48b7vt(GTy%KH0iVd;r|il;@{vro9!RzTWKusG zw-o)`wU3F{%10D>^49W?5g!qXVS&fyNGEBCSww~_q23*v2V2eky^&h#tf7-s)slBD zP%7`EoTYx&@<_$gv3j%-IWV?>s zWK2$$8)4IX6Ei6`0v|5<+3CIzs+^~Dk6?n}Edj@uBEk$wtn1y_Do+a^zwrn^?B?ov zQFjhAiwce>LYAa;iC4RN#>+En-ZyU-RI~+kiHBN+$SbbDRtp%+s&Ma;ufa;IsYaq~5W57zsT^C?}!U~3q- zf|<<0$Ed7Q{B4a@8_7a}{!!?8MZ+5wp2TT7>x8Z?NM-TI>i0nK*zT<%-O6$;`P8$- zZ^-#*crhdJJ#)5D@NBaB$qC2uc_aGT=1Zwf>@kl>ZlqN#tM5$i%D{%$%imzqzyEO; z6{WL82PwC>AfbwJz&_GQDxt*nYMV>3h4$}gN*@$k-0P-t4mte>U=3Y1YJJp6zk{bO z&t1VQJF8OU*S+Ox)YZL&{^}z32h(1`XyLVsur3-UJB$7?$r9E=7nAH3%Lr|xpK*J1 zg1g9jEoRQ$a0**tL}k3QKqg(NIJ!@b)gAO|o@+^?STC#sBQP=5zPKQPybUZ=GsID& zb!{rGzgUbAk(9{+cvyN&02mYs#LKtvs`VJt_2E71Qd_T0jxKha^t+C;~1ebh#4 zI7i?%-&PadP#8ANWGDCZ`kY8mZkJ@w_^*vy)fi*ugd4j6B-sY_=^O9OkWKn470kooi{9DhA}gmPHBam?fbYhD)m}zl$nn2EtQkD zcaF{6w9fpz?*L{`!{#3|^_Rrd@{?2A)fWlxKP0WGtvH&$&(F*lw>;%GoDN%>x8`-$ zS=P(&G6V66vn<;g@;_udjuzk$gch@B$N+^qz%&u&$8t*N_mxxe`CBs8VDcuy682DIl?L!k-t(>c}w7$T&R_Q9ek? zjAp`&+FVXq(uRWdxauTx5fSC2y(G$v8DlB&d8|7*c$wX;o)tB-@5FiK+Y|U#$u*C3 z=RZcI;VsDsXjZCjl`1RHAU?K~S){zh$^>btQ0v(h&68}G&z~qBi6r=zGf~GqszvMo zyp_L~*k}Z&JK#v=r=)nTJD6YQrsC|kuRd={U7T0*a(X^QX7(#HH?M6yIS8N1R;r|b zI`QGnpCM`=N;l6ZQg_JE@X&}fcm+l3YKUsPu`f=Jn0)nHYqDb%eRW{9yoLavg@O1<3-;Vm`7_B>BTJ7~)yvthO^2~4ax7uEcj&iGxYCE%r&dvlneZM~U zvN)mfQMz2qB(M&r52qgOT2!rtdt!@>c=5W1it>ohv4q~}1)0BULl#Kyz~grT`k-@# zeXP6Kf)oSl@=OK%X#7POI4F31;bdU)om52ctpehEGq{IVGeZ?*mmPb1Fm3sYLE=H; zgZL|bg=`GWBlBK?$)=JK*O32R7d-I`!W2MF+c(%ZIBL@8JnHK8zS;U}{KSG+3FW1; z;wGK<&PbrQBD9shv1=e(a|8gZuGW>}0eKaABUu5CburO-T#^(q%DH?7l`$-)GNFK7DjXHd zO;MI33zhptesR~(s|Ux znT>DW3enBPSU!Pz5-UKbx{pc!n=;oA}KOybOesFfE>w;$IZE~Im0BQiOZb~ z+4j>(X@V&p)slG~HIZD-#L_hWaI;^F!iX}6bl~^#rJg4n{O<%A$9}mrwy5ryj4b%8 zCu}3b&-#>Mkv8ROJI=XKbxJCv_S?QbMxTs2 zf$+cQ2S1g2tUgYHxRt_`FPsE}z_q&4sc0}#75|+eKW}18k1u6Jixb38V$vjrXHmew zgC2X6Vo9EI5)r7lEL9CQpsg^~l_t*^47ur`4|`mG@Z3_j7A6bnA?;iqWCqy61D@8;@m4*bMScbAv94ql^hX54{uZ zEAwt|1K?&}kW^DIOYf}=qRP5c_B-2O0|R=gEz_2+CBK7Y7Lr4vYQoL6d-#0awYH%n zBFDPg&i5o!-Ts7=yvOLpU~<7$rb)9#)=j*4smnO+YiNMSV3gu>82fM`%OqaOemW%E zP32ds?6;q>=rlWOM-wOfh5Ol#M)dWs*78HAb@AWLZ_uQK&XF{u zFfjDAq!+LV8ct?5CKh&fdL}kXCMHU1I7WGUQ!yhKGa@Q69u_8ME+%%S|FN`6`~&+2 zRT1W55@ux<{@-IR3pW?h|KL{q z`4N@@B@F*N=PwvEv@Wtjao7Zuz|*dfQ+hAYaBOcAlMB#l`WpVgpVm6a#OjBW z92@GBC;XwrQq`YVkE*=4YA!zatJhXX9IjDUCH+~Yw$$of#qQeAP?gLR2UMD?1 z!?E}psa3KESGxyg_h9-EmE37UHbG+caPgQlnfEC$CWw4NnUi+<4n_da#?8(FM@}xT IAOZKk05ap%MgRZ+ diff --git a/doc/Projects/2017/Project/pdf/Project.tex b/doc/Projects/2017/Project/pdf/Project.tex index 981295dd6..1caed0776 100644 --- a/doc/Projects/2017/Project/pdf/Project.tex +++ b/doc/Projects/2017/Project/pdf/Project.tex @@ -134,8 +134,7 @@ May 2018 \paragraph{Introduction.} The aim of this project is to use an already developed Monte Carlo program for the one-dimensional and two-dimensional \href{{https://github.com/CompPhysics/MachineLearning/tree/master/doc/Programs/IsingModel}}{Ising model}, in order to produce the spin configurations for a series of energies $E_i$ (10000 in total) for a system of $L=40$ spins in one dimension and $L=40\times 40$ in two dimensions at three different temperatures. -In its simplest form -the energy of the Ising model is expressed as, without an externally applied magnetic field, +In its simplest form the energy of the Ising model is expressed as, without an externally applied magnetic field, \[ E=-J\sum_{< kl >}^{N}s_ks_l \] @@ -145,10 +144,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol $$ indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz $J> 0$. We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values $-1$ and $+1$ only. -We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of \href{{https://arxiv.org/abs/1803.08823}}{Mehta et al, arXiv 1803.08823}. This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in \href{{https://physics.bu.edu/~pankajm/MLnotebooks.html}}{this article is accompanied by its jupyter notebook}. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of \href{{https://arxiv.org/abs/1803.08823}}{Mehta et al, arXiv 1803.08823}. This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in \href{{https://physics.bu.edu/~pankajm/MLnotebooks.html}}{this article is accompanied by its jupyter notebook}. This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. @@ -157,15 +156,15 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a \textbf{random forest} +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a \textbf{random forest} algorithm and deep neural networks. You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of \href{{https://compphysics.github.io/MachineLearning/doc/web/course.html}}{FYS-STK4155}. @@ -205,7 +204,7 @@ Use described section 7 and in Notebook 6 of \href{{https://physics.bu.edu/~pank You can use scikit-learn to perform these analyses. See below for instruction on how to install scikit-learn. \paragraph{Part d): Classifying the Ising model phase using neural networks.} -We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta \emph{et al.}). The method is described in \href{{https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html}}{notebook 12}. +We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta \emph{et al.}). The method is described in \href{{https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html}}{notebook 12}. You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow. diff --git a/doc/src/Projects/2017/Project/Project.do.txt b/doc/src/Projects/2017/Project/Project.do.txt index f7573e6e5..3b85ebc07 100644 --- a/doc/src/Projects/2017/Project/Project.do.txt +++ b/doc/src/Projects/2017/Project/Project.do.txt @@ -8,8 +8,7 @@ DATE: May 2018 === Introduction === The aim of this project is to use an already developed Monte Carlo program for the one-dimensional and two-dimensional "Ising model":"https://github.com/CompPhysics/MachineLearning/tree/master/doc/Programs/IsingModel", in order to produce the spin configurations for a series of energies $E_i$ (10000 in total) for a system of $L=40$ spins in one dimension and $L=40\times 40$ in two dimensions at three different temperatures. -In its simplest form -the energy of the Ising model is expressed as, without an externally applied magnetic field, +In its simplest form the energy of the Ising model is expressed as, without an externally applied magnetic field, !bt \[ E=-J\sum_{< kl >}^{N}s_ks_l @@ -21,10 +20,10 @@ constant expressing the strength of the interaction between neighboring spins. The symbol $$ indicates that we sum over nearest neighbors only. We will assume that we have a ferromagnetic ordering, viz $J> 0$. We will use periodic boundary conditions and -the Metropolis algorithm only. +the Metropolis algorithm only. The spins take values $-1$ and $+1$ only. -We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of "Mehta et al, arXiv 1803.08823":"https://arxiv.org/abs/1803.08823". This article stands out as an excellent review of machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in "this article is accompanied by its jupyter notebook":"https://physics.bu.edu/~pankajm/MLnotebooks.html". This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, thir data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. +We will use the Ising model to generate our training data and will focus mainly on supervised training. We will follow closely the recent article of "Mehta et al, arXiv 1803.08823":"https://arxiv.org/abs/1803.08823". This article stands out as an excellent review on machine learning (ML) algorithms applied to typical physics problems. The added benefit is that each figure and model presented in "this article is accompanied by its jupyter notebook":"https://physics.bu.edu/~pankajm/MLnotebooks.html". This means that we can start using these and compare with our own results. In case you wish to use their data for the Ising model, their data can be downloaded from the same link which lists to the jupyter notebooks. See also at the end of the project description for more information on how to install various Python packages. @@ -33,15 +32,15 @@ regression methods, the value of the coupling constant for the energy of the one-dimensional Ising model. Thereafter, we will use the two-dimensional data, but now computed at different temperatures, in order to classify the phase of the Ising model. Below the critical -temperature, the system will be in so-called ferromagnetic -phase. Close to the critical temperature, the final magnetizations -starts becoming close to zero while above the critical temperature, -the net magnetization is zero. The latter case, that is the -two-dimensional Ising model, will be studied using a _random forest_ +temperature, the system will be in a so-called ferromagnetic +phase. Close to the critical temperature, the final magnetization becomes smaller and smaller in absolute value + while above the critical temperature, +the net magnetization is zero. This classification case, that is the +two-dimensional Ising model, will be studied using logistic regression, a _random forest_ algorithm and deep neural networks. You should try to program at least one of these methods yourself (choose the one you prefer). -Feel free to use the notebooks to benchmark your code. If you wish to write your C++ or Fortran program for say a simple neural network model, please feel free to do so. +Feel free to use the notebooks to benchmark your code. If you wish to write your own C++ or Fortran program for say a simple neural network model, please feel free to do so. You can then benchmark your results against the above jupyter notebooks. More information can also be found at the link for the lecture notes of "FYS-STK4155":"https://compphysics.github.io/MachineLearning/doc/web/course.html". @@ -84,7 +83,8 @@ Use described section 7 and in Notebook 6 of "Mehta *et al.*":"https://physics.b You can use scikit-learn to perform these analyses. See below for instruction on how to install scikit-learn. === Part d): Classifying the Ising model phase using neural networks === -We end the classification problem of the phases of the Ising model by emplpying feed-forward deep neural networks (see section 9 of Mehta *et al.*). The method is described in "notebook 12":"https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html". + +We end the classification problem of the phases of the Ising model by employing the algorithm for so-called feed-forward deep neural networks (see section 9 of Mehta *et al.*). The method is described in "notebook 12":"https://physics.bu.edu/~pankajm/ML-Notebooks/HTML/NB_CIX-DNN_ising_TFlow.html". You can use tensorflow to perform these analyses. See below for instruction on how to install tensorflow.