Skip to content

GitLab

Commit 6b974123 authored by vincentvigon's avatar vincentvigon
Browse files

3 premier chap de 'alea'

parent a005fb67
Pipeline #2564 failed with stage
in 2 minutes and 59 seconds
Expand all Hide whitespace changes
Inline Side-by-side
.gitignore
View file @ 6b974123
......@@ -3,4 +3,4 @@
**/__pycache__
**/.ipynb_checkpoints/
**/.DS_Store
**/build
\ No newline at end of file
.idea/inspectionProfiles/Project_Default.xml
View file @ 6b974123
......@@ -5,8 +5,10 @@
<inspection_tool class="PyCompatibilityInspection" enabled="true" level="WARNING" enabled_by_default="true">
<option name="ourVersions">
<value>
<list size="1">
<list size="3">
<item index="0" class="java.lang.String" itemvalue="3.5" />
<item index="1" class="java.lang.String" itemvalue="3.6" />
<item index="2" class="java.lang.String" itemvalue="3.7" />
</list>
</value>
</option>
......
.idea/misc.xml
View file @ 6b974123
......@@ -4,4 +4,7 @@
<option name="languageLevel" value="ES6" />
</component>
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6" project-jdk-type="Python SDK" />
<component name="PythonCompatibilityInspectionAdvertiser">
<option name="version" value="3" />
</component>
</project>
\ No newline at end of file
.idea/workspace.xml
View file @ 6b974123
This diff is collapsed.
alea/00-python_de_base.ipynb
View file @ 6b974123
This diff is collapsed.
alea/01-numpy.ipynb
View file @ 6b974123
This diff is collapsed.
alea/02-histogramme_densite.ipynb deleted 100644 → 0
View file @ a005fb67
This diff is collapsed.
alea/02-plot.ipynb 0 → 100644
View file @ 6b974123
This diff is collapsed.
alea/03-histogramme_densite.ipynb 0 → 100644
View file @ 6b974123
This diff is collapsed.
alea/05-technique_simulation.ipynb 0 → 100644
View file @ 6b974123
This diff is collapsed.
alea/06-TCL_intervalConfiance_student.ipynb
View file @ 6b974123
This diff is collapsed.
alea/07-tests.ipynb 0 → 100644
View file @ 6b974123
This diff is collapsed.
alea/08-Chaine_de_Markov.ipynb
View file @ 6b974123
......@@ -19,302 +19,6 @@
"np.set_printoptions(precision=3,suppress=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Préliminaires d'algèbre linéaire\n",
"\n",
"Attention, en numpy les vecteurs, les matrices lignes et les matrices colonnes sont des objets différents:\n",
"\n",
"* vecteur.shape = (?)\n",
"* matrice_ligne.shape = (1,?)\n",
"* matrice_colonne.shape = (?,1)\n",
"* matrice_quelconque.shape = (?,?)\n",
"\n",
"Observez bien les sorties consoles: les vecteurs s'écrivent avec 1 crochet, les matrices avec 2 crochets.\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Multiplication matricielle\n",
"\n",
"`np.matmul()` s'applique uniquement entre matrices."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"mat_col\n",
" [[1.]\n",
" [1.]\n",
" [1.]]\n",
"\n",
"mat_lin\n",
" [[1. 1. 1.]]\n",
"\n",
"mat_square\n",
" [[1. 1. 1.]\n",
" [1. 1. 1.]\n",
" [1. 1. 1.]]\n",
"\n",
"mat_square × mat_col\n",
" [[3.]\n",
" [3.]\n",
" [3.]]\n",
"\n",
"mat_lin × mat_square\n",
" [[3. 3. 3.]]\n",
"\n",
"mat_lin × mat_square × mat_col\n",
" [[9.]]\n"
]
}
],
"source": [
"size=3\n",
"mat_col=np.ones(shape=[size,1])\n",
"mat_lin=np.ones(shape=[1,size])\n",
"mat_square=np.ones(shape=[size,size])\n",
"\n",
"print(\"\\nmat_col\\n\",mat_col)\n",
"print(\"\\nmat_lin\\n\",mat_lin)\n",
"print(\"\\nmat_square\\n\",mat_square)\n",
"\n",
"print(\"\\nmat_square × mat_col\\n\",np.matmul(mat_square,mat_col) )\n",
"print(\"\\nmat_lin × mat_square\\n\",np.matmul(mat_lin,mat_square))\n",
"print(\"\\nmat_lin × mat_square × mat_col\\n\", np.matmul(np.matmul(mat_lin,mat_square),mat_col))\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"sinon on peut utiliser `np.dot()` qui permet les multiplications matrice × vecteur, matrice × matrice, vecteur × vecteur (=produit scalaire)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"vec0 × vec1\n",
" 6.0\n",
"\n",
"vec0 × mat_square\n",
" [3. 3. 3.]\n",
"\n",
"mat_square × vec0\n",
" [3. 3. 3.]\n",
"\n",
"mat_square × mat_square\n",
" [[3. 3. 3.]\n",
" [3. 3. 3.]\n",
" [3. 3. 3.]]\n"
]
}
],
"source": [
"size=3\n",
"vec0=np.ones(shape=[size])\n",
"vec1=2*np.ones(shape=[size])\n",
"mat_square=np.ones(shape=[size,size])\n",
"\n",
"print(\"\\nvec0 × vec1\\n\",np.dot(vec0,vec1))\n",
"print(\"\\nvec0 × mat_square\\n\",np.dot(vec0,mat_square))\n",
"print(\"\\nmat_square × vec0\\n\",np.dot(mat_square,vec0))\n",
"print(\"\\nmat_square × mat_square\\n\",np.dot(mat_square,mat_square))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"et si vous avez python 3.5+, vous pouver utilisez l'opérateur @ qui rend les codes plus lisibles"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"vec0 × vec1\n",
" 6.0\n",
"\n",
"vec0 × mat_square\n",
" [3. 3. 3.]\n",
"\n",
"mat_square × vec0\n",
" [3. 3. 3.]\n",
"\n",
"mat_square × mat_square\n",
" [[3. 3. 3.]\n",
" [3. 3. 3.]\n",
" [3. 3. 3.]]\n"
]
}
],
"source": [
"print(\"\\nvec0 × vec1\\n\",vec0 @ vec1)\n",
"print(\"\\nvec0 × mat_square\\n\",vec0 @ mat_square)\n",
"print(\"\\nmat_square × vec0\\n\",mat_square @ vec0)\n",
"print(\"\\nmat_square × mat_square\\n\",mat_square @ mat_square)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***Exo:*** Multipliez matriciellement des matrices de taille non-compatible.\n",
"Extrayez la partie intéressante du message d'erreur."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### inverse et transposée \n",
"\n",
"Notez que le pseudo-inverse permet d'inverser les matrices non-inversibles (testez). "
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"mat\n",
" [[1. 0. 2.]\n",
" [0. 1. 0.]\n",
" [0. 0. 1.]]\n",
"\n",
"mat^T\n",
" [[1. 0. 0.]\n",
" [0. 1. 0.]\n",
" [2. 0. 1.]]\n",
"\n",
"mat^(-1)\n",
" [[ 1. 0. -2.]\n",
" [ 0. 1. 0.]\n",
" [ 0. 0. 1.]]\n",
"\n",
"mat^(-1)-pseudo-inverse\n",
" [[ 1. 0. -2.]\n",
" [ 0. 1. 0.]\n",
" [ 0. 0. 1.]]\n"
]
}
],
"source": [
"size=3\n",
"mat=np.zeros(shape=[size,size])\n",
"for i in range(size):\n",
" mat[i,i]=1\n",
"mat[0,size-1]=2\n",
"\n",
"\n",
"print(\"\\nmat\\n\",mat )\n",
"print(\"\\nmat^T\\n\",mat.T)\n",
"print(\"\\nmat^(-1)\\n\",np.linalg.inv(mat))\n",
"print(\"\\nmat^(-1)-pseudo-inverse\\n\",np.linalg.pinv(mat))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### vecteur propre et valeur propre\n",
"\n",
"Souvenez-vous qu'en anglais valeur/vecteur 'propre' c'est 'eigen' value/vector"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"mat\n",
" [[0.085 0.287 0.206 0.422]\n",
" [0.191 0.262 0.253 0.295]\n",
" [0.244 0.07 0.06 0.626]\n",
" [0.383 0.001 0.161 0.454]]\n",
"\n",
"eigVal\n",
" [ 1. 0.122 -0.195 -0.066]\n",
"\n",
"eigVec\n",
" [[-0.5 -0.416 -0.418 0.39 ]\n",
" [-0.5 -0.763 0.307 0.432]\n",
" [-0.5 0.402 -0.739 -0.813]\n",
" [-0.5 0.287 0.43 -0.036]]\n"
]
}
],
"source": [
"size=4\n",
"\"\"\"une matrice aléatoire auquel on fait subir une opération de normalisation (laquelle?)\"\"\"\n",
"mat=np.random.uniform(low=0,high=10,size=[size,size])\n",
"rowSum=np.sum(mat,axis=1)\n",
"\"\"\" mat[i,j]=mat[i,j]/rowSum(i) \"\"\"\n",
"for i in range(size): mat[i,:]/=rowSum[i]\n",
"\n",
"eigVal,eigVec=np.linalg.eig(mat)\n",
"print(\"\\nmat\\n\", mat )\n",
"print(\"\\neigVal\\n\",np.real(eigVal))\n",
"print(\"\\neigVec\\n\", np.real(eigVec))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***A vous:*** \n",
"\n",
"* Pourquoi 1 est-il valeur propre ? Quel est le vecteur propre associé? `np.linalg.eig` donne-t-il les vecteurs propres à droite ou à gauche? Sont-ils disposés en ligne ou bien en colonne?\n",
"\n",
"* Toutes les matrices sont-elles diagonalisables? A quoi servent les np.real()? \n",
"\n",
"* laquelle de ces deux formules est la bonne ?\n",
"\n",
" Diag = P^-1 @ mat @ P \n",
"\n",
"ou\n",
"\n",
" Diag = P @ mat @ P^-1\n",
" \n",
"Vérifiez avec python. Aide: Pour transformez un vecteur en matrice diagonale, utilisez `np.diag(vecteur)`"
]
},
{
"cell_type": "markdown",
"metadata": {},
......
alea/10-RadomVec_GaussianVec.ipynb
View file @ 6b974123
This diff is collapsed.
data_analysis/D_10_LinearModel/linearModel.ipynb → alea/12-linearModel.ipynb
View file @ 6b974123
......@@ -273,7 +273,7 @@
"\n",
"Réponse: on peut faire des stats maintenant! \n",
"* estimer la variance du bruit\n",
"* estimre la variances des estimateurs $\\hat w_i$\n",
"* estimer la variances des estimateurs $\\hat w_i$\n",
"* faire un test pour savoir si l'un des $w_i$ est nul. Ce qui revient à dire que le i-ième input n'a aucune influence sur l'output. Ce genre de considération est importante, car on cherche à expliquer l'output de la manière la plus simple possible. \n",
"\n",
"Au travail:"
......@@ -460,9 +460,9 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.0"
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 1
"nbformat_minor": 2
}
alea/12-GLM.ipynb → alea/13-GLM.ipynb
View file @ 6b974123
This diff is collapsed.
alea/14-permuation_aleatoire.ipynb 0 → 100644
View file @ 6b974123
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Permutation aléatoire"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
alea/boook.tplx 0 → 100644
View file @ 6b974123
% Default to the notebook output style
((* if not cell_style is defined *))
((* set cell_style = 'style_ipython.tplx' *))
((* endif *))
% Inherit from the specified cell style.
((* extends cell_style *))
%===============================================================================
% Latex Book
%===============================================================================
((* block predoc *))
((( super() )))
((* block tableofcontents *))\tableofcontents((* endblock tableofcontents *))
((* endblock predoc *))
((* block docclass *))
\documentclass[9pt, twocolumn, landscape,A4,reprint, floatfix, groupaddress, prb]{book}
\usepackage{placeins}
\AtBeginDocument{
\heavyrulewidth=.08em
\lightrulewidth=.05em
\cmidrulewidth=.03em
\belowrulesep=.65ex
\belowbottomsep=0pt
\aboverulesep=.4ex
\abovetopsep=0pt
\cmidrulesep=\doublerulesep
\cmidrulekern=.5em
\defaultaddspace=.5em
\addtolength{\voffset}{-0.5in}
\addtolength{\hoffset}{-0.5in}
\addtolength{\textheight}{1in}
\addtolength{\textwidth}{1in}
}
%\addtolength{\voffset}{-0.5in}
%\addtolength{\hoffset}{-0.5in}
\usepackage[francais]{babel} % Pour la langue française
\setlength{\itemsep}{0pt}
\setlength{\columnsep}{0.5in}
% Force floating figures to be [H]
\usepackage{float}
\floatplacement{figure}{H}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{mathrsfs}
\usepackage{makeidx}
\usepackage{listings}
\usepackage[svgnames]{xcolor} % Required for colour specification
\definecolor{dkgreen}{rgb}{0,0.6,0}
\definecolor{gray}{rgb}{0.5,0.5,0.5}
\definecolor{mauve}{rgb}{0.58,0,0.82}
\definecolor{verylightgray}{rgb}{0.9,0.9,0.9}
\lstset{frame=tb,
language=Python,
aboveskip=3mm,
belowskip=3mm,
showstringspaces=false,
columns=flexible,
basicstyle={\small\ttfamily},
numbers=none,
numberstyle=\tiny\color{gray},
keywordstyle=\color{blue},
commentstyle=\color{dkgreen},
stringstyle=\color{mauve},
breaklines=true,
breakatwhitespace=false,
tabsize=3,
otherkeywords={self},
backgroundcolor=\color{verylightgray},
}
%\usepackage{showframe}
\definecolor{pythonblue}{rgb}{0.21,0.45,0.65}
\definecolor{pythonyellow}{rgb}{1.,0.77,0.18}
\parindent=0pt
\setlength{\parskip}{1.5ex plus 1.0ex minus 0.5ex}
%\renewcommand{\labelitemi}{$\bullet$}
((* endblock docclass *))
((* block markdowncell scoped *))
((( cell.source | citation2latex | strip_files_prefix | convert_pandoc('markdown', 'json',extra_args=[]) | resolve_references | convert_pandoc('json','latex', extra_args=["--top-level-division=chapter"]) )))
((* endblock markdowncell *))
((* block maketitle *))
\begin{titlepage}
%\vspace*{\baselineskip}
\centering
{\Large Vincent Vigon}
%\vspace*{0.167\textheight}
\raggedleft
\textbf{\fontsize{30}{36}\selectfont Python}\\[\baselineskip]
% \textbf{\textcolor{pythonblue}{\fontsize{50}{60}\selectfont Python}}\\[\baselineskip]
\textcolor{pythonyellow}{\fontsize{30}{36}\selectfont \textit{pour les proba-stats}}
%\vspace*{0.1\textheight}
\centering
\includegraphics{img/champi.jpg}
\vfill
{\large Université de Strasbourg}
\end{titlepage}
((* endblock maketitle *))
((*- block any_cell -*))
((*- if not 'hide' in cell['metadata'].get('tags', {}) -*))
((( super() )))
((*- endif -*))
((*- endblock any_cell -*))
((* block input scoped *))
\begin{lstlisting}[breaklines]
(((cell.source)))
\end{lstlisting}
((* endblock input *))
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment