Outliers_d3.js.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyzing the effect of outliers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This data was generated as part of the following [blog entry](http://jose-coto.com/outlier-visualization). I generated two regressions. The only difference between both is that the second has an outlier. I study graphically the effects of an outlier for a regression. You can check the interactive graph in the blog entry. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import statsmodels.api as sm\n",
    "import matplotlib.pyplot as plt\n",
    "from statsmodels.sandbox.regression.predstd import wls_prediction_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "np.random.seed(seed=5)\n",
    "n_sample = 20\n",
    "indp_var = np.linspace(0, 3, n_sample) \n",
    "indp_var_const = sm.add_constant(indp_var)\n",
    "beta = np.array([0.5, 3])\n",
    "dep_var = np.dot(indp_var_const, beta) + np.random.normal(size=n_sample)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.856\n",
      "Model:                            OLS   Adj. R-squared:                  0.849\n",
      "Method:                 Least Squares   F-statistic:                     107.4\n",
      "Date:                Sat, 27 Aug 2016   Prob (F-statistic):           5.13e-09\n",
      "Time:                        20:27:31   Log-Likelihood:                -29.468\n",
      "No. Observations:                  20   AIC:                             62.94\n",
      "Df Residuals:                      18   BIC:                             64.93\n",
      "Df Model:                           1                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------\n",
      "const          0.7979      0.480      1.663      0.114        -0.210     1.806\n",
      "x1             2.8335      0.273     10.365      0.000         2.259     3.408\n",
      "==============================================================================\n",
      "Omnibus:                        1.697   Durbin-Watson:                   2.403\n",
      "Prob(Omnibus):                  0.428   Jarque-Bera (JB):                1.438\n",
      "Skew:                           0.588   Prob(JB):                        0.487\n",
      "Kurtosis:                       2.417   Cond. No.                         4.24\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "model = sm.OLS(dep_var, indp_var_const)\n",
    "results = model.fit()\n",
    "print(results.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "predicted_values = results.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x111127e48>]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEQCAYAAABPxOQhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nb4uwG6kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e8072b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "plt.style.use('ggplot')\n",
    "fig_1, ax = plt.subplots()\n",
    "ax.scatter(indp_var*3, dep_var)\n",
    "ax.plot(indp_var*3, predicted_values, linewidth=1.5)\n",
    "ax.set_xlim([0, 9])\n",
    "ax.set_title('Regression without outliers', fontsize=15)\n",
    "ax.plot((indp_var*3)[17], dep_var[17],color=\"red\", marker='o', ms=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.311\n",
      "Model:                            OLS   Adj. R-squared:                  0.273\n",
      "Method:                 Least Squares   F-statistic:                     8.126\n",
      "Date:                Sat, 27 Aug 2016   Prob (F-statistic):             0.0106\n",
      "Time:                        20:27:31   Log-Likelihood:                -66.432\n",
      "No. Observations:                  20   AIC:                             136.9\n",
      "Df Residuals:                      18   BIC:                             138.9\n",
      "Df Model:                           1                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------\n",
      "const         -0.8929      3.045     -0.293      0.773        -7.291     5.505\n",
      "x1             4.9470      1.735      2.851      0.011         1.301     8.593\n",
      "==============================================================================\n",
      "Omnibus:                       43.226   Durbin-Watson:                   2.278\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              148.334\n",
      "Skew:                           3.423   Prob(JB):                     6.16e-33\n",
      "Kurtosis:                      14.452   Cond. No.                         4.24\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "#Inserting an outlier and recalculating regression\n",
    "np.random.seed(seed=5)\n",
    "dep_var_2 = np.dot(indp_var_const, beta) + np.random.normal(size=n_sample)\n",
    "dep_var_2[17] = 40\n",
    "model_2 = sm.OLS(dep_var_2, indp_var_const)\n",
    "results = model_2.fit()\n",
    "print(results.summary())\n",
    "predicted_values_2 = results.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1116eb780>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAHPCAYAAAC2ptekAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11114fb00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12,7))\n",
    "ax1.scatter(indp_var*3, dep_var_2)\n",
    "ax1.plot(indp_var*3, predicted_values_2, linewidth=1.5)\n",
    "ax1.plot(indp_var*3, predicted_values, linewidth=1.5)\n",
    "ax1.set_xlim([0, 9])\n",
    "ax1.set_ylim([0, 12])\n",
    "ax2.scatter(indp_var*3, dep_var_2)\n",
    "ax2.plot(indp_var*3, predicted_values_2, linewidth=1.5)\n",
    "ax2.set_xlim([0, 9])\n",
    "ax2.plot((indp_var*3)[17], dep_var_2[17],color=\"red\", marker='o', ms=6)\n",
    "fig.suptitle(\"Cheking the effect of an outlier\", fontsize=14)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "original_ds = np.column_stack((indp_var*3, dep_var, predicted_values, dep_var_2, predicted_values_2))\n",
    "np.savetxt(\"original_ds.csv\", original_ds, delimiter=\",\", fmt= \"%10.5f\", header=\"ind_var,dep_var, pred_1, dep_var_2, pred_val_2 \")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:data_science]",
   "language": "python",
   "name": "conda-env-data_science-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.4.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}