{
 "metadata": {
  "name": "multiple_testing"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "import numpy as np\n",
      "import patsy as pt\n",
      "import statsmodels.api as sm\n",
      "from statsmodels.sandbox.stats.multicomp import multipletests"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "<style>\n",
      "    div.input {\n",
      "        width: 105ex; /* about 80 chars + buffer */\n",
      "    }\n",
      "    div.text_cell {\n",
      "        width: 105ex; /* instead of 100% */\n",
      "    }\n",
      "    div.text_cell_render {\n",
      "        /*font-family: \"Helvetica Neue\", Arial, Helvetica, Geneva, sans-serif;*/\n",
      "        font-family: \"Charis SIL\", serif !important; /* Make non-code text serif. */\n",
      "        line-height: 145% !important; /* added for some line spacing of text. */\n",
      "        width: 105ex !important; /* instead of 'inherit' for shorter lines */\n",
      "    }\n",
      "    div.text_cell_render h1 {\n",
      "        font-size: 18pt;\n",
      "    }\n",
      "    div.text_cell_render h2 {\n",
      "        font-size: 14pt;\n",
      "    }\n",
      "    .CodeMirror {\n",
      "        font-family: Consolas, monospace;\n",
      "    }\n",
      "</style>\n",
      "\n",
      "Test"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "np.random.seed(1010093)\n",
      "\n",
      "pValues = []\n",
      "\n",
      "for i in xrange(1000):\n",
      "    x = np.random.normal(size=20)\n",
      "    y = np.random.normal(size=20)\n",
      "    y, x = pt.dmatrices('y ~ x')\n",
      "\n",
      "    pValues.append(sm.OLS(y, x).fit().pvalues[1])\n",
      "    \n",
      "pValues = np.array(pValues)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "sum(pValues < .05)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 18,
       "text": [
        "54"
       ]
      }
     ],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "_, p_adjust, _, _ = multipletests(pValues, method='bonferroni')\n",
      "sum(p_adjust < .05)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 19,
       "text": [
        "0"
       ]
      }
     ],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "_, p_adjust, _, _ = multipletests(pValues, method='fdr_bh')\n",
      "sum(p_adjust < .05)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 20,
       "text": [
        "0"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "np.random.seed(1010093)\n",
      "\n",
      "pValues = []\n",
      "\n",
      "for i in xrange(1000):\n",
      "    x = np.random.normal(size=20)\n",
      "    \n",
      "    # first 500 beta = 0, last 500 beta = 2\n",
      "    if i < 500:\n",
      "        y = np.random.normal(size=20)\n",
      "    else:\n",
      "        y = np.random.normal(loc=2*x, size=20)\n",
      "\n",
      "    y, x = pt.dmatrices('y ~ x')\n",
      "    \n",
      "    pValues.append(sm.OLS(y, x).fit().pvalues[1])\n",
      "    \n",
      "pValues = np.array(pValues)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "trueStatus = np.concatenate([np.repeat('zero', 500), np.repeat('not zero', 500)])\n",
      "pd.crosstab(pValues < .05, trueStatus)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th>col_0</th>\n",
        "      <th>not zero</th>\n",
        "      <th>zero</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>row_0</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>False</th>\n",
        "      <td>   0</td>\n",
        "      <td> 470</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>True </th>\n",
        "      <td> 500</td>\n",
        "      <td>  30</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 22,
       "text": [
        "col_0  not zero  zero\n",
        "row_0                \n",
        "False         0   470\n",
        "True        500    30"
       ]
      }
     ],
     "prompt_number": 22
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "_, p_adjust, _, _ = multipletests(pValues, method='bonferroni')\n",
      "pd.crosstab(p_adjust < .05, trueStatus)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th>col_0</th>\n",
        "      <th>not zero</th>\n",
        "      <th>zero</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>row_0</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>False</th>\n",
        "      <td>  29</td>\n",
        "      <td> 500</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>True </th>\n",
        "      <td> 471</td>\n",
        "      <td>   0</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 23,
       "text": [
        "col_0  not zero  zero\n",
        "row_0                \n",
        "False        29   500\n",
        "True        471     0"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plot(pValues, p_adjust, 'ok')\n",
      "xlim(-.05, 1.05)\n",
      "ylim(-.05, 1.05)\n",
      "xlabel('pValues')\n",
      "ylabel('p_adjust');"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEICAYAAAC3Y/QeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n"
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "_, p_adjust, _, _ = multipletests(pValues, method='fdr_bh')\n",
      "pd.crosstab(p_adjust < .05, trueStatus)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th>col_0</th>\n",
        "      <th>not zero</th>\n",
        "      <th>zero</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>row_0</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>False</th>\n",
        "      <td>   0</td>\n",
        "      <td> 483</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>True </th>\n",
        "      <td> 500</td>\n",
        "      <td>  17</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 25,
       "text": [
        "col_0  not zero  zero\n",
        "row_0                \n",
        "False         0   483\n",
        "True        500    17"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plot(pValues, p_adjust, 'ok')\n",
      "xlim(-.05, 1.05)\n",
      "ylim(-.05, 1.05)\n",
      "xlabel('pValues')\n",
      "ylabel('p_adjust');"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEICAYAAAC3Y/QeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neSGGPxGRF2L4ExF5IYY/EZEX+n+XKHfV6au9XQAAAABJRU5ErkJggg==\n"
      }
     ],
     "prompt_number": 26
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}