TMaze results preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results Preprocessing for TMaze experiments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "import csv\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=original, step_reward=0 continuous_learning=True'\n",
      "[(0, 5), (5, 22), (10, 27), (15, -1), (20, 25), (25, 3), (30, 3), (35, 27), (40, 11)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=original, step_reward=-0.1 continuous_learning=False'\n",
      "[(0, 11), (5, 96), (10, 46), (15, 97), (20, 96), (25, -1), (30, 56), (35, 67), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=original, step_reward=-0.1 continuous_learning=True'\n",
      "[(0, 5), (5, -1), (10, -1), (15, -1), (20, -1), (25, -1), (30, -1), (35, -1), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=original, step_reward=0 continuous_learning=False'\n",
      "[(0, 15), (5, 23), (10, 21), (15, 21), (20, -1), (25, 40), (30, 50), (35, 50), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=passthru, step_reward=-0.1 continuous_learning=True'\n",
      "[(0, 8), (5, 13), (10, -1), (15, -1), (20, -1), (25, -1), (30, -1), (35, -1), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=passthru, step_reward=0 continuous_learning=True'\n",
      "[(0, 7), (5, 11), (10, 1), (15, 1), (20, 1), (25, 1), (30, 1), (35, 1), (40, 1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=passthru, step_reward=0 continuous_learning=False'\n",
      "[(0, 10), (5, 9), (10, 6), (15, 12), (20, 17), (25, 57), (30, -1), (35, 26), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Experiment 'results/depth_start=0 depth_end=40 agent_type=passthru, step_reward=-0.1 continuous_learning=False'\n",
      "[(0, 11), (5, 15), (10, 8), (15, 28), (20, 25), (25, -1), (30, -1), (35, 143), (40, -1)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Collect results for each experiment, at every depth\n",
    "\n",
    "root = 'results'\n",
    "\n",
    "experiments = {}\n",
    "\n",
    "for experiment_folder in [os.path.join(root, name) for name in os.listdir(root) if os.path.isdir(os.path.join(root, name))]:\n",
    "    \n",
    "    if not os.path.isfile(os.path.join(experiment_folder, \"done.txt\")):\n",
    "        # skip this folder as the experiment has not completed\n",
    "        break\n",
    "    \n",
    "    print(\"Experiment '{}'\".format(experiment_folder))\n",
    "    \n",
    "    depth_iterations = []\n",
    "    \n",
    "    # collect depth runs    \n",
    "    for depth_folder in [os.path.join(experiment_folder, name) for name in os.listdir(experiment_folder) if os.path.isdir(os.path.join(experiment_folder, name))]:\n",
    "        with open(os.path.join(depth_folder, 'results.csv'), 'r') as csvfile:\n",
    "            reader = csv.reader(csvfile)\n",
    "            rows = list(reader)\n",
    "            depth, steps = [int(x) for x in rows[-1]]\n",
    "            depth_iterations.append((depth, steps))\n",
    "            \n",
    "    depth_iterations.sort()        \n",
    "    \n",
    "    print(depth_iterations)\n",
    "    \n",
    "    x,y = zip(*depth_iterations)\n",
    "    \n",
    "    # change non convergence to 0\n",
    "    y = [0 if x == -1 else x for x in y]\n",
    "    \n",
    "    plt.bar(x,y)\n",
    "    \n",
    "    plt.ylabel(\"Number of iterations\")\n",
    "    plt.xlabel(\"Length of corridor\")\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "    with open(os.path.join(experiment_folder, \"all_results.csv\"), 'w') as results:\n",
    "            \n",
    "        writer = csv.writer(results)\n",
    "\n",
    "        head = ['depth','steps']\n",
    "        \n",
    "        writer.writerow(head)\n",
    "        \n",
    "        for row in depth_iterations:\n",
    "            writer.writerow(row)        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def smooth(X, gamma = 0.9):\n",
    "    \"\"\" Returns exponential moving average. \"\"\"\n",
    "    y = np.mean(X[:500])\n",
    "    out = []\n",
    "    for x in X:\n",
    "        y = gamma * y + (1-gamma) * x\n",
    "        out.append(y)\n",
    "    return out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cleanData(filename):\n",
    "    dataframe = pd.read_csv(filename)\n",
    "    X = dataframe.iloc[:,1].values #reward column\n",
    "    print(X)\n",
    "    X = smooth(X,0.99)\n",
    "    dataframe['smoothReward'] = X\n",
    "    #x = dataframe.as_matrix(columns)\n",
    "    #dataframe['smoothReward'] = dataframe['reward'].apply(lambda x: smooth(x,0.99))\n",
    "    dataframe.to_csv(filename, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ -2.82 -23.29 -23.2  ...   3.8    3.8    3.9 ]\n"
     ]
    }
   ],
   "source": [
    "cleanData('rewards.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}