plot_output.ipynb

{
 "cells": [
  {
   "cell_type": "heading",
   "metadata": {
    "collapsed": true
   },
   "level": 1,
   "source": [
    "## Code for visualizing differences between training and development metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib as mpl\n",
    "mpl.use('Agg') # run matplotlib without X server (GUI)\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "log_file = 'out_arctic_attcond_valxtest.txt'\n",
    "\n",
    "metric_to_plot = 'Bleu_4:'\n",
    "id_to_plot = 'Computing coco scores on the'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data = dict()\n",
    "\n",
    "f = open(log_file, 'r')\n",
    "iters_per_val = 1\n",
    "for line in f:\n",
    "    if id_to_plot in line:\n",
    "        line = line.split(' ')[-2].strip()\n",
    "        last_id = line\n",
    "        if last_id not in data.keys():\n",
    "            data[last_id] = []\n",
    "        \n",
    "    if metric_to_plot in line:\n",
    "        line = line.split(' ')[-1].strip()\n",
    "        data[last_id].append(float(line))\n",
    "        \n",
    "    if 'EVAL_EACH_UPDATES' in line:\n",
    "        line = line.split(\"'EVAL_EACH_UPDATES': \")[1]\n",
    "        iters_per_val = int(line.split(\",\")[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[None, None, None, None]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAEQCAYAAACz0c/rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f5639c49490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "positions = range(iters_per_val,len(vals)*iters_per_val+1, iters_per_val)\n",
    "for ids, vals in data.iteritems():\n",
    "    plt.plot(positions, vals)\n",
    "    \n",
    "#plt.ylim([0, 1])\n",
    "plt.xticks(positions)\n",
    "locs, labels = plt.xticks()\n",
    "plt.setp(labels, rotation=90)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2.0
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}