comparison notebooks/sensitivity_experiment.ipynb @ 82:4395037087b6 branch-tests

notebooks

 {
  "cells": [
   {
    "cell_type": "code",
    "execution_count": 1,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "ERROR! Session/line number was not unique in database. History logging moved to new session 32\n"
+     ]
+    },
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
       "/homes/mp305/anaconda/lib/python2.7/site-packages/librosa/core/audio.py:33: UserWarning: Could not import scikits.samplerate. Falling back to scipy.signal\n",

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 2,
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "OUTPUT_FILES = load_dataset.OUTPUT_FILES\n",
     "n_iters = 10"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": 5,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
        "(8396, 108)"

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 48,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 52,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 56,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
        "array([ 59,   1,   1,   1,   1, 733, 733, 733, 733, 733, 733, 733, 733,\n",

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 8,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/html": [
        "<div>\n",

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 47,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 59,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 21,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 52,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 67,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

     "<br> Sort by outlier percentage in descending order."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 68,
+   "execution_count": 7,
    "metadata": {
     "collapsed": true
    },
    "outputs": [],
    "source": [

     "    ranked_outliers = pd.concat([ranked_outliers, df_global['Outliers']], axis=1)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 69,
-   "metadata": {
-    "collapsed": false
-   },
+   "execution_count": 8,
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
        "(137, 10)"
       ]
      },
-     "execution_count": 69,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [

     "Remove countries with 0% outliers as these are in random (probably alphabetical) order."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 70,
-   "metadata": {
-    "collapsed": false
-   },
+   "execution_count": 9,
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    "source": [
     "zero_idx = np.where(np.sum(ranked_outliers, axis=1)==0)[0]\n",
     "first_zero_idx = np.min(zero_idx)\n",
     "ranked_countries = ranked_countries.iloc[:first_zero_idx, :]\n",
     "ranked_outliers = ranked_outliers.iloc[:first_zero_idx, :]\n",
+    "ranked_countries_arr = ranked_countries.get_values()\n",
     "\n",
     "print ranked_countries.head()\n",
     "print ranked_outliers.head()"
    ]
   },

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 72,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 80,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 81,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [

    ]
   },
   {
    "cell_type": "code",
    "execution_count": 76,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
        "{'Chad', 'French Guiana', 'Gambia'}"

    "source": [
     "common_set"
    ]
   },
   {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Try precision at K"
+   ]
+  },
+  {
    "cell_type": "code",
-   "execution_count": 97,
-   "metadata": {
-    "collapsed": true
-   },
+   "execution_count": 10,
+   "metadata": {},
    "outputs": [],
    "source": [
     "# majority voting + precision at K (top5?)\n",
     "from collections import Counter\n",
     "K_vote = 10\n",
     "country_vote = Counter(ranked_countries_arr[:K_vote, :].ravel())"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 98,
-   "metadata": {
-    "collapsed": false
-   },
+   "execution_count": 11,
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/html": [
        "<div>\n",

        "2   Belize   2\n",
        "3     Chad  10\n",
        "4   Bhutan   7"
       ]
      },
-     "execution_count": 98,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [

     "df_country_vote.head()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 99,
-   "metadata": {
-    "collapsed": false
-   },
+   "execution_count": 12,
+   "metadata": {},
    "outputs": [
     {
      "data": {
       "text/html": [
        "<div>\n",

        "0          Brazil   1\n",
        "13           Laos   1\n",
        "9           Malta   1"
       ]
      },
-     "execution_count": 99,
+     "execution_count": 12,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
     "df_country_vote.sort_values(0, ascending=False)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 102,
-   "metadata": {
-    "collapsed": false
-   },
+   "execution_count": 14,
+   "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "0.51000000000000001"
-      ]
-     },
-     "execution_count": 102,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "0.51 0.0830662386292\n"
+     ]
     }
    ],
    "source": [
     "def precision_at_k(array, gr_truth, k):\n",
     "    return len(set(array[:k]) & set(gr_truth[:k])) / float(k)\n",

     "ground_truth = df_country_vote['index'].get_values()\n",
     "p_ = []\n",
     "for j in range(ranked_countries_arr.shape[1]):\n",
     "    p_.append(precision_at_k(ranked_countries_arr[:, j], ground_truth, k))\n",
     "p_ = np.array(p_)\n",
-    "np.mean(p_)"
+    "print np.mean(p_), np.std(p_)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([ 0.6,  0.5,  0.5,  0.5,  0.5,  0.5,  0.5,  0.4,  0.7,  0.4])"
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "p_"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,

    "file_extension": ".py",
    "mimetype": "text/x-python",
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython2",
-   "version": "2.7.11"
+   "version": "2.7.12"
   }
  },
  "nbformat": 4,
  "nbformat_minor": 1
 }