4. ford_compare.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:34:18.680136Z",
     "start_time": "2023-05-18T16:34:18.649147Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from lsr.alternative_ford import bellman_ford, greedy_ford, decay_ford, noisy_ford\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "Notebook to generate Figure 4 from the paper."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:34:46.800138Z",
     "start_time": "2023-05-18T16:34:32.580655Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/pablomonteagudo/anaconda3/envs/nar_env/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "data = np.load('trajs.npz', allow_pickle=True)\n",
    "inputs = data['inputs']\n",
    "score = data['score']\n",
    "lengths = data['lengths']\n",
    "trajs = data['trajs']\n",
    "outputs = data['outputs']\n",
    "hints = data['hints']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:34:52.121882Z",
     "start_time": "2023-05-18T16:34:52.087387Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16384"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n = inputs[0].data.shape[0]\n",
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:34:53.162942Z",
     "start_time": "2023-05-18T16:34:53.126585Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 4, 3, ..., 3, 4, 4], dtype=int32)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lengths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:35:04.610099Z",
     "start_time": "2023-05-18T16:35:02.558598Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "all_hints = np.array([x['pi_h'] for x in hints[:9]])[:,lengths==9,...]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:35:05.817198Z",
     "start_time": "2023-05-18T16:35:05.486199Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 0, 64)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "allpreds = np.argmax(all_hints, axis=-1)\n",
    "allpreds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:35:07.288517Z",
     "start_time": "2023-05-18T16:35:07.251516Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 64)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "narpreds = allpreds[-1]\n",
    "narpreds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:35:09.654512Z",
     "start_time": "2023-05-18T16:35:08.971513Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "allAs = np.array([inputs[2].data[i, ...] for i in range(n)])[lengths==9]\n",
    "allss = np.array([np.argmax(inputs[1].data[i, ...]) for i in range(n)])[lengths==9]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:35:59.409962Z",
     "start_time": "2023-05-18T16:35:10.399966Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "bf_out = [bellman_ford(allAs[i], allss[i], 9) for i in range(allss.shape[0])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:37:44.924628Z",
     "start_time": "2023-05-18T16:37:44.893598Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0,)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_preds = np.array([x[0] for x in bf_out])\n",
    "all_preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:37:45.171180Z",
     "start_time": "2023-05-18T16:37:45.142145Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "too many indices for array: array is 1-dimensional, but 3 were indexed",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[12], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m true_preds \u001b[38;5;241m=\u001b[39m \u001b[43mall_preds\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m:\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m      2\u001b[0m true_preds\u001b[38;5;241m.\u001b[39mshape\n",
      "\u001b[0;31mIndexError\u001b[0m: too many indices for array: array is 1-dimensional, but 3 were indexed"
     ]
    }
   ],
   "source": [
    "true_preds = all_preds[:,-1,:]\n",
    "true_preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:38:37.919681Z",
     "start_time": "2023-05-18T16:37:45.367181Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "decay_out = np.array([decay_ford(allAs[i], allss[i], 9) for i in range(allss.shape[0])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:38:37.935682Z",
     "start_time": "2023-05-18T16:38:37.921680Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4047, 64)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "decay_preds = decay_out[:,-1,:]\n",
    "decay_preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:39:27.643180Z",
     "start_time": "2023-05-18T16:38:37.936181Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "greedy_out =  np.array([greedy_ford(allAs[i], allss[i], 9) for i in range(allss.shape[0])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:39:27.658681Z",
     "start_time": "2023-05-18T16:39:27.645182Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4047, 64)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "greedy_preds = greedy_out[:,-1,:]\n",
    "greedy_preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:42:05.432680Z",
     "start_time": "2023-05-18T16:39:27.660180Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "noisy_out =  np.array([noisy_ford(allAs[i], allss[i], 9) for i in range(allss.shape[0])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:42:05.448180Z",
     "start_time": "2023-05-18T16:42:05.432680Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4047, 64)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "noisy_preds = noisy_out[:,-1,:]\n",
    "noisy_preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-18T16:55:29.606024Z",
     "start_time": "2023-05-18T16:55:29.064524Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x275 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(5,2.75))\n",
    "\n",
    "#ax1\n",
    "\n",
    "x = [narpreds, true_preds, noisy_preds, decay_preds, greedy_preds]\n",
    "total = narpreds.reshape(-1, 1).shape[0]\n",
    "conf = np.array([[np.count_nonzero(x[i]==x[j]) / total for i in range(5)] for j in range(5)])\n",
    "labels = ['NAR', 'Bellman', 'Noisy', 'Decay', 'Greedy']\n",
    "\n",
    "sns.heatmap(conf, annot=True, fmt='.0%', cbar=False, cmap='Blues', xticklabels=labels, yticklabels=labels, ax=ax1)\n",
    "\n",
    "#ax2\n",
    "\n",
    "was_true = np.any(allpreds == true_preds, axis=0)\n",
    "is_true = narpreds == true_preds\n",
    "change = np.array([[np.count_nonzero(was_true & is_true), np.count_nonzero((~was_true) & is_true)],\n",
    "          [np.count_nonzero(was_true & (~is_true)), np.count_nonzero((~was_true) & (~is_true))]]) / total\n",
    "labels = ['Correct', 'Incorrect']\n",
    "\n",
    "sns.heatmap(change, annot=True, fmt='.2%', cbar=False, cmap='Blues',xticklabels=labels, yticklabels=labels, ax=ax2)\n",
    "\n",
    "ax2.set_xlabel('Intermediate Predictions')\n",
    "ax2.set_ylabel('Final Predictions')\n",
    "\n",
    "#\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig('plots_new/4.png', dpi=400)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}