Auswertung_200er.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import glob\n",
      "import csv\n",
      "from operator import add\n",
      "import scipy.stats as stat\n",
      "\n",
      "#Dateiliste erstellen\n",
      "files = []\n",
      "for filename in glob.glob('/home/weiss/data/*.csv'):\n",
      "    files.append(filename)\n",
      "    \n",
      "#read files\n",
      "seen = 0\n",
      "trials = 0\n",
      "catchseen = 0\n",
      "catchtrials = 0\n",
      "deltas = []\n",
      "freqs = []\n",
      "for filename in files:\n",
      "    data = csv.reader(open(filename, \"rb\"), delimiter=',')\n",
      "    data.next()\n",
      "    data.next()\n",
      "    data.next()\n",
      "    for row in data:\n",
      "        freqdiff = abs(int(row[0][0:-2])-int(row[1][0:-2]))\n",
      "        if float(row[3]) < float(row[4]):\n",
      "            if row[0] != row[1]:\n",
      "                trials += 1\n",
      "                if row[5] == 'y':\n",
      "                    seen += 1\n",
      "                    deltas.append((freqdiff,1))\n",
      "                    freqs.append((int(row[0][0:-2]),int(row[1][0:-2]),1))\n",
      "                else:\n",
      "                    deltas.append((freqdiff,0))\n",
      "                    freqs.append((int(row[0][0:-2]),int(row[1][0:-2]),0))\n",
      "            if row[0] == row[1]:\n",
      "                catchtrials += 1\n",
      "                if row[5] == 'n':\n",
      "                    catchseen += 1\n",
      "        else:\n",
      "            print row\n",
      "                                        \n",
      "d_delta = {}\n",
      "for freq, count in deltas:\n",
      "    if freq not in d_delta:\n",
      "        d_delta[freq] = [0,0]\n",
      "    d_delta[freq][0] += 1\n",
      "    if count:\n",
      "        d_delta[freq][1] += 1\n",
      "        \n",
      "d_freqs = {}\n",
      "for freq1, freq2, count in freqs:\n",
      "    if (freq1, freq2) not in d_freqs:\n",
      "        d_freqs[(freq1, freq2)] = [0,0]\n",
      "    d_freqs[(freq1, freq2)][0] += 1\n",
      "    if count:\n",
      "        d_freqs[(freq1, freq2)][1] += 1\n",
      "        \n",
      "\n",
      "d_all = d_freqs.copy()\n",
      "d_all['overall\\ncorrectness'] = (1, round(float(seen)/trials, 3))\n",
      "d_all['false\\npositives'] = (1, 1-round(float(catchseen)/catchtrials, 3))\n",
      "\n",
      "print 'seeing-prob:', round(100.*seen/trials, 3), '%'\n",
      "print 'catchtrial accuracy:', round(100.*catchseen/catchtrials), '%'\n",
      "\n",
      "delta_std = loadtxt(\"/home/weiss/data/delta_std.txt\", delimiter=', ')\n",
      "d_all_std = loadtxt(\"/home/weiss/data/d_all_std.txt\", delimiter=', ')\n",
      "contrast_std = loadtxt(\"/home/weiss/data/contrast_std.txt\", delimiter=', ')\n",
      "\n",
      "print delta_std[:,2].shape\n",
      "\n",
      "delta_std = [stat.sem(delta_std[:,i]) for i in range(delta_std.shape[1])]\n",
      "d_all_std = [stat.sem(d_all_std[:,i]) for i in range(d_all_std.shape[1])]\n",
      "contrast_std = [stat.sem(contrast_std[:,i]) for i in range(contrast_std.shape[1])]\n",
      "\n",
      "f, p = plt.subplots(1, figsize=(12,7))\n",
      "\n",
      "ids = [2, 6, 8, 15, 17, 21]\n",
      "zweihunderter = [(200,400), (400,200), (400,600), (600,400), (600,800), (800,600)]\n",
      "\n",
      "zweihundert = [(200,400), (400,200), (200,500), (500,200), (200,600), (600,200), (200,800), (800,200)]\n",
      "ids2 = [2,6,3,10,4,14,5,18]\n",
      "\n",
      "niedrige = []\n",
      "ids3 = []\n",
      "\n",
      "p.bar(range(6), [float(d_all[key][1])/d_all[key][0] for key in zweihunderter], align='center', alpha=.3, width=.2, yerr=[d_all_std[i] for i in ids], ecolor='k', error_kw={'elinewidth':2})\n",
      "p.set_xticks(range(6))\n",
      "p.set_xticklabels(zweihunderter)\n",
      "p.set_ylim(0, 1.1)\n",
      "p.set_xlim(-1, 6)\n",
      "#p.grid()\n",
      "#p.set_title('Probability of Seeing', fontsize=15)\n",
      "p.set_xlabel('Frequenzen [Hz]', fontsize=15)\n",
      "p.set_ylabel('P', fontsize=15)\n",
      "plt.xticks(fontsize=15)\n",
      "plt.yticks(fontsize=15)\n",
      "plt.tight_layout()\n",
      "#plt.savefig('/home/weiss/ba/figures/delta200.pdf')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "['200.0', '800.0', '1.3', '0.6', '0.30572104454', 'y', '1']\n",
        "['400.0', '800.0', '1.4', '1.1', '1.01087594032', 'n', '0']\n",
        "seeing-prob: 52.504 %\n",
        "catchtrial accuracy: 95.0 %\n",
        "(6,)\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAA1oAAAHzCAYAAAA5LmyRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f72946c2310>"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}