{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Analyse de risque : Navette Challenger" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "D'après l'article servant de base à cet exercice, les auteurs estiment plusieurs paramètres : $s_{\\hat{\\alpha}} = 3.052$ et $s_{\\hat{\\beta}} = 0.047$. La qualité de l'ajustement est caractérisée par un coefficient $G^2 = 18.086$ avec 21 degrés de liberté." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Informations techniques sur la machine et l'installation python" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Importation des librairies" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "%matplotlib inline\n", "import sys\n", "import platform\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "import statsmodels.api as sm\n", "import seaborn as sn" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Détermination de la version des différentes librairies ainsi que de l'OS" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "3.6.4 |Anaconda, Inc.| (default, Mar 13 2018, 01:15:57) \n", "[GCC 7.2.0]\n", "uname_result(system='Linux', node='d8fc5e21ebcb', release='4.4.0-164-generic', version='#192-Ubuntu SMP Fri Sep 13 12:02:50 UTC 2019', machine='x86_64', processor='x86_64')\n", "IPython \t 7.12.0\n", "IPython.core.release \t 7.12.0\n", "PIL \t 7.0.0\n", "PIL.Image \t 7.0.0\n", "PIL._version \t 7.0.0\n", "_csv \t 1.0\n", "_ctypes \t 1.1.0\n", "_curses \t b'2.2'\n", "decimal \t 1.70\n", "argparse \t 1.1\n", "backcall \t 0.1.0\n", "cffi \t 1.13.2\n", "csv \t 1.0\n", "ctypes \t 1.1.0\n", "cycler \t 0.10.0\n", "dateutil \t 2.8.1\n", "decimal \t 1.70\n", "decorator \t 4.4.1\n", "distutils \t 3.6.4\n", "ipaddress \t 1.0\n", "ipykernel \t 5.1.4\n", "ipykernel._version \t 5.1.4\n", "ipython_genutils \t 0.2.0\n", "ipython_genutils._version \t 0.2.0\n", "ipywidgets \t 7.2.1\n", "ipywidgets._version \t 7.2.1\n", "jedi \t 0.16.0\n", "json \t 2.0.9\n", "jupyter_client \t 6.0.0\n", "jupyter_client._version \t 6.0.0\n", "jupyter_core \t 4.6.3\n", "jupyter_core.version \t 4.6.3\n", "kiwisolver \t 1.1.0\n", "logging \t 0.5.1.2\n", "matplotlib \t 2.2.3\n", "matplotlib.backends.backend_agg \t 2.2.3\n", "numpy \t 1.15.2\n", "numpy.core \t 1.15.2\n", "numpy.core.multiarray \t 3.1\n", "numpy.lib \t 1.15.2\n", "numpy.linalg._umath_linalg \t b'0.1.5'\n", "numpy.matlib \t 1.15.2\n", "optparse \t 1.5.3\n", "pandas \t 0.22.0\n", "_libjson \t 1.33\n", "parso \t 0.6.0\n", "patsy \t 0.5.1\n", "patsy.version \t 0.5.1\n", "pexpect \t 4.8.0\n", "pickleshare \t 0.7.5\n", "platform \t 1.0.8\n", "prompt_toolkit \t 3.0.3\n", "ptyprocess \t 0.6.0\n", "pygments \t 2.5.2\n", "pyparsing \t 2.4.6\n", "pytz \t 2019.3\n", "re \t 2.2.1\n", "scipy \t 1.1.0\n", "scipy._lib.decorator \t 4.0.5\n", "scipy._lib.six \t 1.2.0\n", "scipy.fftpack._fftpack \t b'$Revision: $'\n", "scipy.fftpack.convolve \t b'$Revision: $'\n", "scipy.integrate._dop \t b'$Revision: $'\n", "scipy.integrate._ode \t $Id$\n", "scipy.integrate._odepack \t 1.9 \n", "scipy.integrate._quadpack \t 1.13 \n", "scipy.integrate.lsoda \t b'$Revision: $'\n", "scipy.integrate.vode \t b'$Revision: $'\n", "scipy.interpolate._fitpack \t 1.7 \n", "scipy.interpolate.dfitpack \t b'$Revision: $'\n", "scipy.linalg \t 0.4.9\n", "scipy.linalg._fblas \t b'$Revision: $'\n", "scipy.linalg._flapack \t b'$Revision: $'\n", "scipy.linalg._flinalg \t b'$Revision: $'\n", "scipy.ndimage \t 2.0\n", "scipy.optimize._cobyla \t b'$Revision: $'\n", "scipy.optimize._lbfgsb \t b'$Revision: $'\n", "scipy.optimize._minpack \t 1.10 \n", "scipy.optimize._nnls \t b'$Revision: $'\n", "scipy.optimize._slsqp \t b'$Revision: $'\n", "scipy.optimize.minpack2 \t b'$Revision: $'\n", "scipy.signal.spline \t 0.2\n", "scipy.sparse.linalg.eigen.arpack._arpack \t b'$Revision: $'\n", "scipy.sparse.linalg.isolve._iterative \t b'$Revision: $'\n", "scipy.special.specfun \t b'$Revision: $'\n", "scipy.stats.mvn \t b'$Revision: $'\n", "scipy.stats.statlib \t b'$Revision: $'\n", "seaborn \t 0.8.1\n", "seaborn.external.husl \t 2.1.0\n", "seaborn.external.six \t 1.10.0\n", "six \t 1.14.0\n", "statsmodels \t 0.9.0\n", "statsmodels.__init__ \t 0.9.0\n", "traitlets \t 4.3.3\n", "traitlets._version \t 4.3.3\n", "urllib.request \t 3.6\n", "zlib \t 1.0\n", "zmq \t 17.1.2\n", "zmq.sugar \t 17.1.2\n", "zmq.sugar.version \t 17.1.2\n" ] } ], "source": [ "print(sys.version)\n", "print(platform.uname())\n", "\n", "for name, val in sorted(sys.modules.items()):\n", " try:\n", " print(val.__name__, \"\\t\", val.__version__)\n", " except Exception as e:\n", " pass" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Début de l'étude" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Chargement des données" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Nous commençons par charger les données provenant du lien [https://app-learninglab.inria.fr/moocrr/gitlab/moocrr-session3/moocrr-reproducibility-study/raw/master/data/shuttle.csv](https://app-learninglab.inria.fr/moocrr/gitlab/moocrr-session3/moocrr-reproducibility-study/raw/master/data/shuttle.csv). __Soulignons que le lien donné par [l'exemple](https://app-learninglab.inria.fr/moocrr/gitlab/moocrr-session3/moocrr-reproducibility-study/blob/master/src/Python3/challenger.ipynb) ne peut pas être lu correctement par la librairie pandas__." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Date Count Temperature Pressure Malfunction\n", "0 4/12/81 6 66 50 0\n", "1 11/12/81 6 70 50 1\n", "2 3/22/82 6 69 50 0\n", "3 11/11/82 6 68 50 0\n", "4 4/04/83 6 67 50 0\n", "5 6/18/82 6 72 50 0\n", "6 8/30/83 6 73 100 0\n", "7 11/28/83 6 70 100 0\n", "8 2/03/84 6 57 200 1\n", "9 4/06/84 6 63 200 1\n", "10 8/30/84 6 70 200 1\n", "11 10/05/84 6 78 200 0\n", "12 11/08/84 6 67 200 0\n", "13 1/24/85 6 53 200 2\n", "14 4/12/85 6 67 200 0\n", "15 4/29/85 6 75 200 0\n", "16 6/17/85 6 70 200 0\n", "17 7/2903/85 6 81 200 0\n", "18 8/27/85 6 76 200 0\n", "19 10/03/85 6 79 200 0\n", "20 10/30/85 6 75 200 2\n", "21 11/26/85 6 76 200 0\n", "22 1/12/86 6 58 200 1\n" ] } ], "source": [ "d = pd.read_csv(\"https://app-learninglab.inria.fr/moocrr/gitlab/moocrr-session3/moocrr-reproducibility-study/raw/master/data/shuttle.csv\")\n", "print(d)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Visualisation graphique des données" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Le fichier ne possédant aucune ligne _nulle_ , nous pouvons continuer en toute tranquillité. Nous voulons obtenir un aperçu graphique du nombre d'accidents comme une fonction de la température ambiante." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEKCAYAAAAB0GKPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAHLBJREFUeJzt3X+QXWWd5/H3pzs/SeLAJNJqAkM0lE6UEKENIKidUVhwdshYGRfc0VgqxigZC8YZobZ2LMdxpwbUsdZdNGaQWbEWMyhGU2sw4Ye9DAqahIkJMER7A5ImkpgYJA1J6E5/949zGm+a/vHc7nvur/68qrr6nuc8z+nv956+/e3znHvPUURgZmY2mpZaB2BmZo3BBcPMzJK4YJiZWRIXDDMzS+KCYWZmSVwwzMwsiQuGmZklccEwM7MkLhhmZpZkUq0DqKQ5c+bEGWecUeswTvDcc88xY8aMWodRcc6r8TRrbs5rfLZt23YgIl6e0repCsYZZ5zB1q1bax3GCTo7O+no6Kh1GBXnvBpPs+bmvMZH0i9T+3pKyszMkhRaMCRdKmmXpC5J1w+xfpmkHZK2S9oq6aKSdU9I2jmwrsg4zcxsdIVNSUlqBW4CLga6gS2SNkTEoyXd7gE2RERIWgTcDryuZP3SiDhQVIxmZpauyCOMJUBXROyOiBeAdcCy0g4R0RO/u776DMDXWjczq1NFFoy5wJ6S5e687QSS3iXpMeD7wAdLVgWwWdI2SSsLjNPMzBKoqBsoSXo38B8i4qp8+X3Akoj4i2H6vxX4VES8I19+VUTslXQqcBfwFxFx3xDjVgIrAdra2s5dt25dIfmMVU9PDzNnzqx1GBXnvBpPs+bmvMZn6dKl2yKiPaVvkW+r7QZOK1meB+wdrnNE3CfpNZLmRMSBiNibt++XtJ5siuslBSMi1gJrAdrb26Pe3l7nt/w1lmbNC5o3N+dVPUVOSW0BzpQ0X9IU4EpgQ2kHSQskKX98DjAFOChphqRZefsM4BLg4QJjNTOzURR2hBERfZJWA5uAVuCWiHhE0qp8/RpgObBCUi9wBLgif8dUG7A+ryWTgNsi4gdFxWpmZqMr9JPeEbER2DiobU3J4xuAG4YYtxs4u8jYzMysPP6kt5mZJXHBAA72HONne57hYM+xWodiZmU42HOMI73H/dqtkglfML63/SkuvOFe3nvzT7jwhnvZsP2pWodkZgkGXruP//o5v3arZEIXjIM9x7jujh0c7e3n8LE+jvb288k7dvi/FbM6V/raPR7h126VTOiC0X3oCJNbTnwKJre00H3oSI0iMrMUfu3WxoQuGPNOmU5vf/8Jbb39/cw7ZXqNIjKzFH7t1saELhizZ07lxuWLmDa5hVlTJzFtcgs3Ll/E7JlTax2amY2g9LXbKvm1WyVNdce9sbh88VwuXDCH7kNHmHfKdP/CmTWIgdfuTx+4nx9dfpFfu1Uw4QsGZP+t+JfNrPHMnjmV6ZNb/fqtkgk9JWVmZulcMMzMLIkLhpmZJXHBMDOzJC4YZmaWxAXDzMySuGCYmVkSFwwzM0vigmFmZklcMMzMLEmhBUPSpZJ2SeqSdP0Q65dJ2iFpu6Stki5KHWtmZtVVWMGQ1ArcBFwGLATeI2nhoG73AGdHxGLgg8DNZYw1M7MqKvIIYwnQFRG7I+IFYB2wrLRDRPREROSLM4BIHWtmZtVVZMGYC+wpWe7O204g6V2SHgO+T3aUkTzWzMyqp8jLm2uItnhJQ8R6YL2ktwJ/B7wjdSyApJXASoC2tjY6OzvHGm8henp66i6mSnBejadZc3Ne1VNkwegGTitZngfsHa5zRNwn6TWS5pQzNiLWAmsB2tvbo6OjY5xhV1ZnZyf1FlMlOK/G06y5Oa/qKXJKagtwpqT5kqYAVwIbSjtIWiBJ+eNzgCnAwZSxZmZWXYUdYUREn6TVwCagFbglIh6RtCpfvwZYDqyQ1AscAa7IT4IPObaoWM3MbHSF3qI1IjYCGwe1rSl5fANwQ+pYMzOrHX/S28zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlsQFw8zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlsQFw8zMkrhgmJlZEhcMMzNLUmjBkHSppF2SuiRdP8T6P5e0I//6saSzS9Y9IWmnpO2SthYZp5mZjW5SURuW1ArcBFwMdANbJG2IiEdLuj0OvC0iDkm6DFgLnFeyfmlEHCgqRjMzS1fkEcYSoCsidkfEC8A6YFlph4j4cUQcyhcfBOYVGI+ZmY1DkQVjLrCnZLk7bxvOh4A7S5YD2Cxpm6SVBcRnZmZlKGxKCtAQbTFkR2kpWcG4qKT5wojYK+lU4C5Jj0XEfUOMXQmsBGhra6Ozs3PcgVdST09P3cVUCc6r8TRrbs6reoosGN3AaSXL84C9gztJWgTcDFwWEQcH2iNib/59v6T1ZFNcLykYEbGW7NwH7e3t0dHRUcEUxq+zs5N6i6kSnFfjadbcnFf1FDkltQU4U9J8SVOAK4ENpR0knQ58B3hfRPy8pH2GpFkDj4FLgIcLjNXMzEZR2BFGRPRJWg1sAlqBWyLiEUmr8vVrgE8Bs4EvSwLoi4h2oA1Yn7dNAm6LiB8UFauZmY2uyCkpImIjsHFQ25qSx1cBVw0xbjdw9uB2MzOrHX/S28zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlsQFw8zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlsQFw8zMkiQVDEltkr4m6c58eaGkDxUbmpmZ1ZPUI4z/RXZv7lflyz8HrikiIDMzq0+pBWNORNwO9ANERB9wfLRBki6VtEtSl6Trh1j/55J25F8/lnR26lgzM6uu1ILxnKTZQABIOh/47UgDJLUCNwGXAQuB90haOKjb48DbImIR8HfA2jLGmplZFU1K7PeXwAbgNZJ+BLwc+LNRxiwBuiJiN4CkdcAy4NGBDhHx45L+DwLzUseamVl1JRWMiHhI0tuA1wICdkVE7yjD5gJ7Spa7gfNG6P8h4M5yx0paCawEaGtro7Ozc5Swqqunp6fuYqoE59V4mjU351U9SQVD0tXA/46IR/LlUyS9JyK+PNKwIdpimO0vJSsYF5U7NiLWkk9ltbe3R0dHxwghVV9nZyf1FlMlOK/G06y5Oa/qST2H8eGIeGZgISIOAR8eZUw3cFrJ8jxg7+BOkhYBNwPLIuJgOWPNzKx6UgtGi6QX/+vPT0pPGWXMFuBMSfMlTQGuJDsP8iJJpwPfAd4XET8vZ6yZmVVX6knvTcDtktaQTQ2tAn4w0oCI6JO0Oh/bCtwSEY9IWpWvXwN8CpgNfDmvR30R0T7c2PLTMzOzSkktGNcBHwE+SnZ+YTPZNNKIImIjsHFQ25qSx1cBV6WONTOz2kl9l1Q/8JX8y8zMJqDUd0ldCHwa+IN8jICIiFcXF5qZmdWT1CmprwHXAttIuCSImZk1n9SC8duIuHP0bmZm1qxSC8YPJX2O7C2wxwYaI+KhQqIyM7O6k1owBi7L0V7SFsAfVTYcMzOrV6nvklpadCBmZlbffMc9MzNL4jvumZlZkkLvuGdmZs2jsDvuWTEO9hzjZ3ue4WDPsYr2bTTNnFsRuvYd5tDzvXTtO1zrUKyBFXnHPauw721/iuvu2MHklhZ6+/u5cfkiLl88d9x9G00z51aET313J7c++CSfOKuPa794HysuOJ3PLDur1mFZAxr1CENSCzANeBvwZrKLEL4+InYUHJuVONhzjOvu2MHR3n4OH+vjaG8/n7xjx5D/YZfTt9E0c25F6Np3mFsffPKEtlsfeNJHGjYmoxaM/MKDX4iIvoh4JCIeTrg9q1VY96EjTG45cXdNbmmh+9CRcfVtNM2cWxG273mmrHazkaSew9gsaXnpTZSsuuadMp3e/v4T2nr7+5l3yvRx9W00zZxbERafdnJZ7WYjSS0Yfwl8Czgm6VlJhyU9W2BcNsjsmVO5cfkipk1uYdbUSUyb3MKNyxcxe+bUcfVtNM2cWxEWtM1ixQWnn9C24oLTWdA2q0YRWSNL/aS3f7vqwOWL53Lhgjl0HzrCvFOmj/hHspy+jaaZcyvCZ5adxYrzz2Dntge5+9rzXSxszFLvh/HWodoj4r7KhmOjmT1zavIfyHL6Nppmzq0IC9pm0X3SZBcLG5fUt9X+dcnjacASsntj+OKDZmYTRNI5jIj4k5Kvi4E3APtGGyfpUkm7JHVJun6I9a+T9ICkY5L+atC6JyTtlLRd0tbUhMzMrBipRxiDdZMVjWFJagVuAi7O+2+RtCEiHi3p9hvg48CfDrOZpRFxYIwxmplZBaWew/gf5JcFITsqWQz8bJRhS4CuiNidb2MdsAx4sWBExH5gv6Q/LjNuMzOrstQjjNIpoT7gmxHxo1HGzAX2lCx387sbMaUIss9/BPDViFg7VCdJK4GVAG1tbXR2dpbxI4rX09NTdzFVgvNqPM2am/OqntSC8W3gaEQch2y6SdJJEfH8CGOG+pBfDNE2nAsjYq+kU4G7JD021Luy8kKyFqC9vT06OjrK+BHF6+zspN5iqgTn1XiaNTfnVT2pH9y7Byj9KO104O5RxnQDp5UszwP2pgYWEXvz7/uB9WRTXGZmViOpBWNaRPQMLOSPTxplzBbgTEnzJU0BriS74u2oJM2QNGvgMXAJ8HBirGZmVoDUKannJJ0TEQ8BSDoXGPFqbxHRJ2k12Z36WoFbIuIRSavy9WskvYLs/MjLgH5J1wALgTnA+vzSVZOA2yLiB+WnZ2ZmlZJaMK4BviVpYErplcAVow2KiI3AxkFta0oeP002VTXYs8DZibGZmVkVpF5Laouk1wGvJTuZ/ZgvcW5mNrEkncOQdDUwI78Xxk5gpqSPFRuamZnVk9ST3h+OiBfvuBIRh4APFxOSmZnVo9SC0VJ686T8sh9TignJzMzqUepJ783A7ZLWkH347qOA37VkZjaBpBaMvyGbglpFdtJ7M/C1ooIyM7P6M2LBkDQJ+HvgA2TXhRLZp7cfJ5vOOl50gGZmVh9GO4fxOeD3gVdHxDkR8UZgPvB7wOeLDs7MzOrHaAXjP5K9Q+rwQEP++KPAO4sMzMzM6stoBSMi4iVXmM2vWlvOlWfNzKzBjVYwHpW0YnCjpPcCjxUTkpmZ1aPR3iV1NfAdSR8EtpEdVbyJ7PLm7yo4NjMzqyMjFoyIeAo4T9IfAa8ne5fUnRFxTzWCMzOz+pF68cF7gXsLjsXMzOpY6qVBzMxsgnPBMDOzJC4YZmaWxAXDzMySuGCYmVmSQguGpEsl7ZLUJen6Ida/TtIDko5J+qtyxpqZWXUVVjDymyzdBFwGLATeI2nhoG6/AT7OoAsZJo41M7MqKvIIYwnQFRG7I+IFYB2wrLRDROyPiC1Ab7ljzcysulJvoDQWc8nuoTGgGziv0mMlrQRWArS1tdHZ2Vl2oEXq6empu5gqwXk1nmbNzXlVT5EFQ0O0pV7hNnlsRKwF1gK0t7dHR0dH4o+ojs7OTuotpkpwXo2nWXNzXtVT5JRUN9nd+QbMA/ZWYayZmRWgyIKxBThT0nxJU4ArgQ1VGGtmZgUobEoqIvokrQY2Aa3ALRHxiKRV+fo1kl4BbAVeBvRLugZYGBHPDjW2qFjNzGx0RZ7DICI2AhsHta0pefw02XRT0lgzM6sdf9LbzMySuGCYmVkSFwwzM0vigmFmZklcMMzMLIkLhpmZJXHBMDOzJC4YZmaWxAXDzMySuGCYmVkSFwwzM0vigmFmZklcMMzMLIkLhpmZJXHBMDOzJC4YZmaWxAXDzMySuGCYmVkSFwwzM0tSaMGQdKmkXZK6JF0/xHpJ+lK+foekc0rWPSFpp6TtkrYWGaeZmY1uUlEbltQK3ARcDHQDWyRtiIhHS7pdBpyZf50HfCX/PmBpRBwoKkYzM0tX5BHGEqArInZHxAvAOmDZoD7LgFsj8yBwsqRXFhiTmZmNUWFHGMBcYE/JcjcnHj0M12cu8CsggM2SAvhqRKwd6odIWgmsBGhra6Ozs7MiwVdKT09P3cVUCc6r8TRrbs6reoosGBqiLcroc2FE7JV0KnCXpMci4r6XdM4KyVqA9vb26OjoGEfIldfZ2Um9xVQJzqvxNGtuzqt6ipyS6gZOK1meB+xN7RMRA9/3A+vJprjMzKxGiiwYW4AzJc2XNAW4EtgwqM8GYEX+bqnzgd9GxK8kzZA0C0DSDOAS4OECYzUzs1EUNiUVEX2SVgObgFbgloh4RNKqfP0aYCPwTqALeB74QD68DVgvaSDG2yLiB0XFamZmoyvyHAYRsZGsKJS2rSl5HMDVQ4zbDZxdZGxmZlYef9LbzMySuGCYmVkSFwwzM0vigmFmZklcMMzMLIkLhpmZJXHBMDOzJC4YZmaWxAXDzMySuGCYmVkSFwwzM0vigmFmZklcMMzMLIkLhpmZJXHBMDOzJC4YZmaWxAXDzMySuGCYmVkSFwwzM0tSaMGQdKmkXZK6JF0/xHpJ+lK+foekc1LHmtXSwZ5j/GzPMxzsOTZq362PH+QfN+9i6+MHK7rdcvp27TvMoed76dp3eNS+qYqKtdwYjvQeT34Ovr11T1M+B0VsdyiTitqwpFbgJuBioBvYImlDRDxa0u0y4Mz86zzgK8B5iWPNauJ725/iujt2MLmlhd7+fm5cvojLF88dsu97b36Q+7uyQvGle7t4y4LZfOOq88e93XL6fuq7O7n1wSf5xFl9XPvF+1hxwel8ZtlZY8i8+FjHEsPH/7CXa2+4N+k5GNBsz0GltzucIo8wlgBdEbE7Il4A1gHLBvVZBtwamQeBkyW9MnGsWdUd7DnGdXfs4GhvP4eP9XG0t59P3rFjyP/utj5+8MViMeBfuw4OeaRRznbL6du17/AJfygBbn3gyXH9l11UrGON4XjEhH8OKrndkSgiitmw9GfApRFxVb78PuC8iFhd0uf/AP8QEffny/cA1wFnjDa2ZBsrgZUAbW1t565bt66QfMaqp6eHmTNn1jqMipuoeR3pPc7jv36O4yWvm1aJ+S+fwfTJrSf03ffsMfYfPvqSbZw6axptL5s65u2W0/fQ8710H3oegLbpsO9I1j7vlJM45aTJw+Y5kqJiHWsMA3mlPAel6v05qOTv4kiWLl26LSLaU/oWNiUFaIi2wdVpuD4pY7PGiLXAWoD29vbo6OgoI8TidXZ2Um8xVcJEzetgzzGuveFejvb2v9g2bXILP7r8ImbPPLEIbH38INd99cGXbOPbH2mnff7sMW+3nL5d+w5z7RfvA+ATZ/XxhZ3ZS/7ua89nQdusYfMcSVGxjjWGgbxSnoNS9f4cVPJ3sVKKnJLqBk4rWZ4H7E3skzLWrOpmz5zKjcsXMW1yC7OmZn+kbly+aMgXaPv82bxlwYmF4S0LZr+kWJS73XL6LmibxYoLTj+hbcUFp4/5D2WRsY41hlZpwj8HldzuiCKikC+yo5fdwHxgCvAz4PWD+vwxcCfZEcX5wE9Txw71de6550a9+eEPf1jrEAox0fM6cPhobH/yUBw4fHTUvlt2H4gvbHostuw+UNHtltP3F08/G9/5/ub4xdPPjto3VVGxlhvDxs13Jz8H39ryZMM8B0X8Lg4F2BqJf9cLm5KKiD5Jq4FNQCtwS0Q8ImlVvn4NsBF4J9AFPA98YKSxRcVqVq7ZM6cm/yfXPn/oo4rxbrecvgvaZtF90uRx/Vc9np9fTt9yY5g+uTVp2wvaZlU0/4GfXw/PQaFHFSWKPIdBRGwkKwqlbWtKHgdwdepYMzOrHX/S28zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlsQFw8zMkrhgmJlZksLuh1ELkn4N/LLWcQwyBzhQ6yAK4LwaT7Pm5rzG5w8i4uUpHZuqYNQjSVsj8eYkjcR5NZ5mzc15VY+npMzMLIkLhpmZJXHBKN7aWgdQEOfVeJo1N+dVJT6HYWZmSXyEYWZmSVwwKkjSE5J2StouaWve9mlJT+Vt2yW9s9ZxlkvSyZK+LekxSf8u6QJJvy/pLkm/yL+fUus4x2KY3Bp6n0l6bUns2yU9K+maRt9nI+TV0PsLQNK1kh6R9LCkb0qaVo/7y1NSFSTpCaA9Ig6UtH0a6ImIz9cqrvGS9HXgXyPiZklTgJOA/wL8JiL+QdL1wCkRcV1NAx2DYXK7hgbfZwMktQJPAeeR3Q654fcZvCSvD9DA+0vSXOB+YGFEHJF0O9ntqRdSZ/vLRxg2IkkvA94KfA0gIl6IiGeAZcDX825fB/60NhGO3Qi5NZO3A/8vIn5JE+yzEqV5NYNJwHRJk8j+adlLHe4vF4zKCmCzpG2SVpa0r5a0Q9It9XBYWaZXA78G/lnSv0m6WdIMoC0ifgWQfz+1lkGO0XC5QWPvs1JXAt/MHzfDPhtQmhc08P6KiKeAzwNPAr8CfhsRm6nD/eWCUVkXRsQ5wGXA1ZLeCnwFeA2wmOyX4Qs1jG8sJgHnAF+JiDcCzwHX1zakihkut0bfZwDkU2yXA9+qdSyVNEReDb2/8gK3DJgPvAqYIem9tY1qaC4YFRQRe/Pv+4H1wJKI2BcRxyOiH/gnYEktYxyDbqA7In6SL3+b7I/sPkmvBMi/769RfOMxZG5NsM8GXAY8FBH78uVm2GcwKK8m2F/vAB6PiF9HRC/wHeDN1OH+csGoEEkzJM0aeAxcAjw8sMNz7wIerkV8YxURTwN7JL02b3o78CiwAXh/3vZ+4Hs1CG9chsut0fdZifdw4rRNw++z3Al5NcH+ehI4X9JJkkT2e/jv1OH+8rukKkTSq8mOKiCb6rgtIv6bpG+QHSoH8ATwkYF5yUYhaTFwMzAF2E32rpQW4HbgdLJf+HdHxG9qFuQYDZPbl2j8fXYSsAd4dUT8Nm+bTYPvs2HyaobX2N8CVwB9wL8BVwEzqbP95YJhZmZJPCVlZmZJXDDMzCyJC4aZmSVxwTAzsyQuGGZmlmRSrQMwK1r+dtJ78sVXAMfJLgkC2YcrX6hJYCOQ9EFgY/5ZEbO64LfV2oRST1cPltQaEceHWXc/sDoitpexvUkR0VexAM0G8ZSUTWiS3i/pp/l9FL4sqUXSJEnPSPqcpIckbZJ0nqT/K2n3wP0WJF0laX2+fpek/5q43c9K+imwRNLfStqS3wdhjTJXkH0Q7V/y8VMkdUs6Od/2+ZLuzh9/VtJXJd1FdhHFSZL+Mf/ZOyRdVf1n1ZqVC4ZNWJLeQHYpiTdHxGKyKdor89W/B2zOLyb5AvBpsks2vBv4TMlmluRjzgH+s6TFCdt9KCKWRMQDwH+PiDcBZ+XrLo2IfwG2A1dExOKEKbM3An8SEe8DVgL7I2IJ8Cayi2CePpbnx2wwn8OwiewdZH9Ut2aX8GE62WUnAI5ExF35451kl5zuk7QTOKNkG5si4hCApO8CF5G9robb7gv87hIyAG+X9NfANGAOsA24s8w8vhcRR/PHlwB/KKm0QJ1JdmkJs3FxwbCJTMAtEfE3JzRmN7Ep/a++HzhW8rj0dTP4JGCMst0jkZ84zK+L9D/JrpD7lKTPkhWOofTxuxmBwX2eG5TTxyLiHswqzFNSNpHdDfwnSXMgezfVGKZvLlF2X/CTyO5p8KMytjudrAAdyK90vLxk3WFgVsnyE8C5+ePSfoNtAj6WF6eB+2BPLzMnsyH5CMMmrIjYmV8l9G5JLUAvsIrs9pip7gduI7uBzzcG3tWUst2IOKjsnuIPA78EflKy+p+BmyUdITtP8mngnyQ9Dfx0hHi+SnZ10+35dNh+skJmNm5+W63ZGOXvQHpDRFxT61jMqsFTUmZmlsRHGGZmlsRHGGZmlsQFw8zMkrhgmJlZEhcMMzNL4oJhZmZJXDDMzCzJ/wep3cGhQHbwqAAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "d['Occurence'] = d.Malfunction / d.Count\n", "d.plot(x='Temperature', y='Occurence', kind='scatter')\n", "plt.grid(True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Analyse statistique des risques" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Nous cherchons maintenant à déterminer la probabilité qu'un joint soit détruit. Pour cela, nous utilisons un outil de régression logistique disponible dans la librairie statsmodels." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Generalized Linear Model Regression Results
Dep. Variable: Occurence No. Observations: 23
Model: GLM Df Residuals: 21
Model Family: Binomial Df Model: 1
Link Function: logit Scale: 1.0000
Method: IRLS Log-Likelihood: -23.526
Date: Thu, 24 Sep 2020 Deviance: 18.086
Time: 12:55:04 Pearson chi2: 30.0
No. Iterations: 6 Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err z P>|z| [0.025 0.975]
Intercept 5.0850 3.052 1.666 0.096 -0.898 11.068
Temperature -0.1156 0.047 -2.458 0.014 -0.208 -0.023
" ], "text/plain": [ "\n", "\"\"\"\n", " Generalized Linear Model Regression Results \n", "==============================================================================\n", "Dep. Variable: Occurence No. Observations: 23\n", "Model: GLM Df Residuals: 21\n", "Model Family: Binomial Df Model: 1\n", "Link Function: logit Scale: 1.0000\n", "Method: IRLS Log-Likelihood: -23.526\n", "Date: Thu, 24 Sep 2020 Deviance: 18.086\n", "Time: 12:55:04 Pearson chi2: 30.0\n", "No. Iterations: 6 Covariance Type: nonrobust\n", "===============================================================================\n", " coef std err z P>|z| [0.025 0.975]\n", "-------------------------------------------------------------------------------\n", "Intercept 5.0850 3.052 1.666 0.096 -0.898 11.068\n", "Temperature -0.1156 0.047 -2.458 0.014 -0.208 -0.023\n", "===============================================================================\n", "\"\"\"" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "d['ok'] = d.Count - d.Malfunction\n", "d['Intercept'] = 1\n", "\n", "LogisticModel = sm.GLM(d['Occurence'], d[['Intercept','Temperature']], family=sm.families.Binomial(sm.families.links.logit()), var_weights=d['Count']).fit()\n", "LogisticModel.summary()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Nous obtenons $G^2 = 18.086$ ainsi que les erreurs standards sur les coefficients $s_\\hat{\\alpha} = 3.052$ et $s_\\hat{\\beta} = 0.047$. Les résultats de l'article sont donc correctement reproductibles." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Probabilité d'accident" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Nous utilisons le modèle logisitique afin de prévoir la probabilité d'occurrence d'un accident en fonction de la température initiale." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Intercept Temperature Occurence\n", "0 1 30.0 0.834373\n", "1 1 31.0 0.817774\n", "2 1 32.0 0.799911\n", "3 1 33.0 0.780766\n", "4 1 34.0 0.760339\n", "5 1 35.0 0.738645\n", "6 1 36.0 0.715721\n", "7 1 37.0 0.691626\n", "8 1 38.0 0.666441\n", "9 1 39.0 0.640269\n", "10 1 40.0 0.613235\n", "11 1 41.0 0.585485\n", "12 1 42.0 0.557181\n", "13 1 43.0 0.528501\n", "14 1 44.0 0.499631\n", "15 1 45.0 0.470765\n", "16 1 46.0 0.442092\n", "17 1 47.0 0.413800\n", "18 1 48.0 0.386066\n", "19 1 49.0 0.359052\n", "20 1 50.0 0.332904\n", "21 1 51.0 0.307745\n", "22 1 52.0 0.283679\n", "23 1 53.0 0.260787\n", "24 1 54.0 0.239124\n", "25 1 55.0 0.218729\n", "26 1 56.0 0.199617\n", "27 1 57.0 0.181787\n", "28 1 58.0 0.165220\n", "29 1 59.0 0.149886\n", ".. ... ... ...\n", "31 1 61.0 0.122744\n", "32 1 62.0 0.110830\n", "33 1 63.0 0.099940\n", "34 1 64.0 0.090011\n", "35 1 65.0 0.080981\n", "36 1 66.0 0.072783\n", "37 1 67.0 0.065357\n", "38 1 68.0 0.058640\n", "39 1 69.0 0.052575\n", "40 1 70.0 0.047106\n", "41 1 71.0 0.042180\n", "42 1 72.0 0.037749\n", "43 1 73.0 0.033767\n", "44 1 74.0 0.030192\n", "45 1 75.0 0.026985\n", "46 1 76.0 0.024110\n", "47 1 77.0 0.021535\n", "48 1 78.0 0.019229\n", "49 1 79.0 0.017166\n", "50 1 80.0 0.015321\n", "51 1 81.0 0.013671\n", "52 1 82.0 0.012197\n", "53 1 83.0 0.010880\n", "54 1 84.0 0.009703\n", "55 1 85.0 0.008653\n", "56 1 86.0 0.007716\n", "57 1 87.0 0.006879\n", "58 1 88.0 0.006133\n", "59 1 89.0 0.005467\n", "60 1 90.0 0.004873\n", "\n", "[61 rows x 3 columns]\n" ] } ], "source": [ "dLogistic = pd.DataFrame({'Temperature': np.linspace(start=30, stop=90, num=61), 'Intercept': 1})\n", "sm.add_constant(dLogistic)\n", "dLogistic['Occurence'] = LogisticModel.predict(dLogistic)\n", "print(dLogistic)" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "dLogistic.plot(x='Temperature', y='Occurence', kind='line')\n", "plt.scatter(x=d['Temperature'], y=d['Occurence'])\n", "plt.grid(True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Bien que l'utilisation d'un tel modèle pour prédire la probabilité d'un accident soit hautement discutable (trop peu de points et plusieurs points pathologiques), il en résulte sûrement un résultat surestimé. Ainsi, pour une valeur de $T = 31$ °F, la probabilité d'accident serait très importante." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Zone de confiance" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Nous cherchons maintenant à évaluer l'incertitude pour le modèle logistique en fonction de la valeur en température envisagée." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/opt/conda/lib/python3.6/site-packages/scipy/stats/stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.\n", " return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sn.set(color_codes=True)\n", "plt.xlim(30,90)\n", "plt.ylim(0,1)\n", "sn.regplot(x='Temperature', y='Occurence', data=d, logistic=True)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Il apparaît ainsi que malgré l'incertitude, il y ait au moins une probabilité de 10% d'accident." ] }, { "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": "3.6.4" } }, "nbformat": 4, "nbformat_minor": 4 }