fig1plot.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "from load_data import load_data\n",
    "import torch\n",
    "from utils import get_all_walks, plot_mutagenicity\n",
    "from tqdm import tqdm\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_graphs: 4337\n"
     ]
    }
   ],
   "source": [
    "graphs, pos_idx, neg_idx = load_data('Mutagenicity')\n",
    "nn = torch.load(\"models/gin-3-mutagenicity.torch\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "idx = 22\n",
    "order = 3\n",
    "\n",
    "\n",
    "model_depth = len(nn.blocks)-1\n",
    "gammas = np.linspace(3,0,model_depth)\n",
    "\n",
    "g = graphs[idx]\n",
    "node_candidates = np.arange(g.nbnodes)\n",
    "shape = (len(node_candidates),)*(model_depth +1)\n",
    "rel_tensor = torch.zeros(*shape)\n",
    "walks = get_all_walks(2, g.get_adj())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "label is 0\n",
      "order:  3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "ename": "FileNotFoundError",
     "evalue": "[Errno 2] No such file or directory: 'imgs/mol_mut_lohohist.svg'",
     "output_type": "error",
     "traceback": [
      "\u001B[1;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[1;31mFileNotFoundError\u001B[0m                         Traceback (most recent call last)",
      "\u001B[1;32m~\\AppData\\Local\\Temp\\ipykernel_9624\\2720921127.py\u001B[0m in \u001B[0;36m<module>\u001B[1;34m\u001B[0m\n\u001B[0;32m     10\u001B[0m         \u001B[1;32mpass\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m     11\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m---> 12\u001B[1;33m \u001B[0mplot_mutagenicity\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mg\u001B[0m\u001B[1;33m,\u001B[0m\u001B[0msub_relevances\u001B[0m\u001B[1;33m,\u001B[0m\u001B[0mfactor\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;36m1\u001B[0m\u001B[1;33m,\u001B[0m\u001B[0mfigname\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m'imgs/mol_mut_lohohist.svg'\u001B[0m\u001B[1;33m,\u001B[0m\u001B[0mcolor\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m'grey'\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m",
      "\u001B[1;32m~\\Documents\\AVT\\GitHub\\sgnn_lrp_via_mp\\utils.py\u001B[0m in \u001B[0;36mplot_mutagenicity\u001B[1;34m(g, relevances, width, shrinking_factor, factor, color, figname, dataset)\u001B[0m\n\u001B[0;32m    343\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    344\u001B[0m     \u001B[1;32mif\u001B[0m \u001B[0mfigname\u001B[0m \u001B[1;32mis\u001B[0m \u001B[1;32mnot\u001B[0m \u001B[1;32mNone\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 345\u001B[1;33m         \u001B[0mplt\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0msavefig\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfigname\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mdpi\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;36m600\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mformat\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m'eps'\u001B[0m\u001B[1;33m,\u001B[0m\u001B[0mbbox_inches\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m'tight'\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\pyplot.py\u001B[0m in \u001B[0;36msavefig\u001B[1;34m(*args, **kwargs)\u001B[0m\n\u001B[0;32m    956\u001B[0m \u001B[1;32mdef\u001B[0m \u001B[0msavefig\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0margs\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    957\u001B[0m     \u001B[0mfig\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mgcf\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 958\u001B[1;33m     \u001B[0mres\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mfig\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0msavefig\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0margs\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    959\u001B[0m     \u001B[0mfig\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mcanvas\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mdraw_idle\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m)\u001B[0m   \u001B[1;31m# need this if 'transparent=True' to reset colors\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    960\u001B[0m     \u001B[1;32mreturn\u001B[0m \u001B[0mres\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\figure.py\u001B[0m in \u001B[0;36msavefig\u001B[1;34m(self, fname, transparent, **kwargs)\u001B[0m\n\u001B[0;32m   3010\u001B[0m                         ax.patch._cm_set(facecolor='none', edgecolor='none'))\n\u001B[0;32m   3011\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m-> 3012\u001B[1;33m             \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mcanvas\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mprint_figure\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfname\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m   3013\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   3014\u001B[0m     def ginput(self, n=1, timeout=30, show_clicks=True,\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\backend_bases.py\u001B[0m in \u001B[0;36mprint_figure\u001B[1;34m(self, filename, dpi, facecolor, edgecolor, orientation, format, bbox_inches, pad_inches, bbox_extra_artists, backend, **kwargs)\u001B[0m\n\u001B[0;32m   2318\u001B[0m                         \u001B[0morientation\u001B[0m\u001B[1;33m=\u001B[0m\u001B[0morientation\u001B[0m\u001B[1;33m,\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   2319\u001B[0m                         \u001B[0mbbox_inches_restore\u001B[0m\u001B[1;33m=\u001B[0m\u001B[0m_bbox_inches_restore\u001B[0m\u001B[1;33m,\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m-> 2320\u001B[1;33m                         **kwargs)\n\u001B[0m\u001B[0;32m   2321\u001B[0m             \u001B[1;32mfinally\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   2322\u001B[0m                 \u001B[1;32mif\u001B[0m \u001B[0mbbox_inches\u001B[0m \u001B[1;32mand\u001B[0m \u001B[0mrestore_bbox\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\_api\\deprecation.py\u001B[0m in \u001B[0;36mwrapper\u001B[1;34m(*inner_args, **inner_kwargs)\u001B[0m\n\u001B[0;32m    410\u001B[0m                          \u001B[1;32melse\u001B[0m \u001B[0mdeprecation_addendum\u001B[0m\u001B[1;33m,\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    411\u001B[0m                 **kwargs)\n\u001B[1;32m--> 412\u001B[1;33m         \u001B[1;32mreturn\u001B[0m \u001B[0mfunc\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0minner_args\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0minner_kwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    413\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    414\u001B[0m     \u001B[0mDECORATORS\u001B[0m\u001B[1;33m[\u001B[0m\u001B[0mwrapper\u001B[0m\u001B[1;33m]\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mdecorator\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\backends\\backend_ps.py\u001B[0m in \u001B[0;36mprint_eps\u001B[1;34m(self, outfile, *args, **kwargs)\u001B[0m\n\u001B[0;32m    845\u001B[0m     \u001B[1;33m@\u001B[0m\u001B[0m_api\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mdelete_parameter\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;34m\"3.5\"\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;34m\"args\"\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    846\u001B[0m     \u001B[1;32mdef\u001B[0m \u001B[0mprint_eps\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mself\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0moutfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m*\u001B[0m\u001B[0margs\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 847\u001B[1;33m         \u001B[1;32mreturn\u001B[0m \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0m_print_ps\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0moutfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;34m'eps'\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    848\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    849\u001B[0m     \u001B[1;33m@\u001B[0m\u001B[0m_api\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mdelete_parameter\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;34m\"3.4\"\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;34m\"dpi\"\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\_api\\deprecation.py\u001B[0m in \u001B[0;36mwrapper\u001B[1;34m(*inner_args, **inner_kwargs)\u001B[0m\n\u001B[0;32m    384\u001B[0m             \u001B[1;31m# Early return in the simple, non-deprecated case (much faster than\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    385\u001B[0m             \u001B[1;31m# calling bind()).\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 386\u001B[1;33m             \u001B[1;32mreturn\u001B[0m \u001B[0mfunc\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0minner_args\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0minner_kwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    387\u001B[0m         \u001B[0marguments\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0msignature\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mbind\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0minner_args\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0minner_kwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0marguments\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    388\u001B[0m         \u001B[1;32mif\u001B[0m \u001B[0mis_varargs\u001B[0m \u001B[1;32mand\u001B[0m \u001B[0marguments\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mget\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mname\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\backends\\backend_ps.py\u001B[0m in \u001B[0;36m_print_ps\u001B[1;34m(self, outfile, format, dpi, metadata, papertype, orientation, **kwargs)\u001B[0m\n\u001B[0;32m    887\u001B[0m                    self._print_figure)\n\u001B[0;32m    888\u001B[0m         printer(outfile, format, dpi=dpi, dsc_comments=dsc_comments,\n\u001B[1;32m--> 889\u001B[1;33m                 orientation=orientation, papertype=papertype, **kwargs)\n\u001B[0m\u001B[0;32m    890\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    891\u001B[0m     \u001B[1;33m@\u001B[0m\u001B[0m_check_savefig_extra_args\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\backend_bases.py\u001B[0m in \u001B[0;36mwrapper\u001B[1;34m(*args, **kwargs)\u001B[0m\n\u001B[0;32m   1641\u001B[0m             \u001B[0mkwargs\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mpop\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0marg\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1642\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m-> 1643\u001B[1;33m         \u001B[1;32mreturn\u001B[0m \u001B[0mfunc\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m*\u001B[0m\u001B[0margs\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;33m**\u001B[0m\u001B[0mkwargs\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m   1644\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1645\u001B[0m     \u001B[1;32mreturn\u001B[0m \u001B[0mwrapper\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\backends\\backend_ps.py\u001B[0m in \u001B[0;36m_print_figure\u001B[1;34m(self, outfile, format, dpi, dsc_comments, orientation, papertype, bbox_inches_restore)\u001B[0m\n\u001B[0;32m   1022\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1023\u001B[0m         \u001B[1;32melse\u001B[0m\u001B[1;33m:\u001B[0m  \u001B[1;31m# Write directly to outfile.\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m-> 1024\u001B[1;33m             \u001B[1;32mwith\u001B[0m \u001B[0mcbook\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mopen_file_cm\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0moutfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;34m\"w\"\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mencoding\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m\"latin-1\"\u001B[0m\u001B[1;33m)\u001B[0m \u001B[1;32mas\u001B[0m \u001B[0mfile\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m   1025\u001B[0m                 \u001B[1;32mif\u001B[0m \u001B[1;32mnot\u001B[0m \u001B[0mfile_requires_unicode\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1026\u001B[0m                     \u001B[0mfile\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mcodecs\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mgetwriter\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;34m\"latin-1\"\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\cbook\\__init__.py\u001B[0m in \u001B[0;36mopen_file_cm\u001B[1;34m(path_or_file, mode, encoding)\u001B[0m\n\u001B[0;32m    463\u001B[0m \u001B[1;32mdef\u001B[0m \u001B[0mopen_file_cm\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mpath_or_file\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mmode\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;34m\"r\"\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mencoding\u001B[0m\u001B[1;33m=\u001B[0m\u001B[1;32mNone\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    464\u001B[0m     \u001B[1;34mr\"\"\"Pass through file objects and context-manage path-likes.\"\"\"\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 465\u001B[1;33m     \u001B[0mfh\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mopened\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mto_filehandle\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mpath_or_file\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mmode\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;32mTrue\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mencoding\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    466\u001B[0m     \u001B[1;32mreturn\u001B[0m \u001B[0mfh\u001B[0m \u001B[1;32mif\u001B[0m \u001B[0mopened\u001B[0m \u001B[1;32melse\u001B[0m \u001B[0mcontextlib\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mnullcontext\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfh\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    467\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32m~\\miniconda3\\envs\\PyTorch37\\lib\\site-packages\\matplotlib\\cbook\\__init__.py\u001B[0m in \u001B[0;36mto_filehandle\u001B[1;34m(fname, flag, return_opened, encoding)\u001B[0m\n\u001B[0;32m    449\u001B[0m             \u001B[0mfh\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mbz2\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mBZ2File\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfname\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mflag\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    450\u001B[0m         \u001B[1;32melse\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 451\u001B[1;33m             \u001B[0mfh\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mopen\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfname\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mflag\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mencoding\u001B[0m\u001B[1;33m=\u001B[0m\u001B[0mencoding\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    452\u001B[0m         \u001B[0mopened\u001B[0m \u001B[1;33m=\u001B[0m \u001B[1;32mTrue\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    453\u001B[0m     \u001B[1;32melif\u001B[0m \u001B[0mhasattr\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfname\u001B[0m\u001B[1;33m,\u001B[0m \u001B[1;34m'seek'\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;31mFileNotFoundError\u001B[0m: [Errno 2] No such file or directory: 'imgs/mol_mut_lohohist.svg'"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x333.36 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "subgraph = [0,1,3,5,7,9,11,14,15]\n",
    "    \n",
    "print('label is', g.label)\n",
    "print('order: ', order)\n",
    "sub_relevances = []\n",
    "for walk in walks:\n",
    "    a,b= walk\n",
    "    if a in subgraph and b in subgraph:\n",
    "        sub_relevances.append([walk, .4 ])\n",
    "        pass\n",
    "\n",
    "plot_mutagenicity(g,sub_relevances,factor=1,figname='imgs/mol_mut_lohohist.svg',color='grey')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analyzing the distribution of subgraph relevances when considering different attribution orders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7/7 [01:27<00:00, 12.47s/it]\n"
     ]
    }
   ],
   "source": [
    "import subgraph_relevance as sr\n",
    "from itertools import combinations\n",
    "node_label_dict = {0:'C',1:'O',2:'Cl',3:'H',4:'N',5:'F',6:'Br',7:'S',8:'P',9:'I',10:'Na',11:'K',12:'Li',13:'Ca'}\n",
    "sample_ids = range(2000) \n",
    "#######\n",
    "\n",
    "all_subfgs = []\n",
    "atoms = ['C', 'N','O']\n",
    "for size in range(1,len(atoms)+1): all_subfgs += list(combinations(atoms, size))\n",
    "model_depth = len(nn.blocks)-1\n",
    "\n",
    "gammas = list(np.linspace(3, 0, model_depth))\n",
    "\n",
    "lo_rels = {}\n",
    "ho_rels = {}\n",
    "for subgraph_by_atoms in tqdm(all_subfgs):\n",
    "    los, hos = [], []\n",
    "    for idx in sample_ids:\n",
    "        g = graphs[idx]\n",
    "        node_candidates = np.arange(g.nbnodes)\n",
    "        # shape = (len(node_candidates),)*(model_depth +1)\n",
    "        atom_numb_list = [node_label_dict[node_id] for node_id in g.node_tags]\n",
    "        S = [idx for idx in node_candidates if atom_numb_list[idx] in subgraph_by_atoms]\n",
    "\n",
    "        ho_subrel = sr.subgraph_mp_forward_hook(nn, g, S, gamma=gammas)\n",
    "        lo_subrel = nn.lrp(g.adj,gammas,g.label,None,H0=g.node_features)[S].sum()\n",
    "        los.append(lo_subrel.numpy()); hos.append(ho_subrel.numpy())\n",
    "        \n",
    "    ho_rels[subgraph_by_atoms] = np.array(hos)\n",
    "    lo_rels[subgraph_by_atoms] = np.array(los)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 324x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 324x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from functools import reduce\n",
    "import matplotlib.patches as mpatches\n",
    "\n",
    "subfgs = all_subfgs #[7:14]\n",
    "samples = [22]\n",
    "cmap = plt.get_cmap('viridis')\n",
    "\n",
    "def interaction_rel(fg, rels, only_atoms=False):\n",
    "    if len(fg) == 1: \n",
    "        return rels[fg]\n",
    "    if len(fg) == 2: \n",
    "        first_order_fgs = [(at,) for at in fg]\n",
    "        outrel = rels[fg].copy()\n",
    "        for subst_fg in first_order_fgs: \n",
    "            outrel -= rels[subst_fg]\n",
    "        return outrel\n",
    "    if len(fg) == 3:\n",
    "        first_order_fgs = [(at,) for at in fg]\n",
    "        second_order_fgs = list(combinations(list(fg), 2)) \n",
    "        \n",
    "        outrel = rels[fg].copy()\n",
    "        for subst_fg in first_order_fgs: \n",
    "            outrel += rels[subst_fg]\n",
    "        \n",
    "        for subst_fg in second_order_fgs: \n",
    "            outrel -= rels[subst_fg]\n",
    "            \n",
    "        return outrel\n",
    "\n",
    "for order,rels in zip(['lo','ho'],[lo_rels,ho_rels]):\n",
    "    plt.figure(figsize=(4.5,5))\n",
    "    plt.rcParams.update({'font.size': 18})\n",
    "    ax = plt.subplot(1,1,1)\n",
    "    names = []\n",
    "    xs=range(len(subfgs))\n",
    "    color_patches=[]\n",
    "    \n",
    "    for fg,color,x in zip(subfgs, torch.linspace(0,1,len(subfgs)), xs):\n",
    "        name = reduce(lambda a,b: a+b, fg)\n",
    "        names.append(name)\n",
    "        if order == 'lo':\n",
    "            color=(0.7,.45,.0,1)\n",
    "        else:\n",
    "            color=(0.,.25,.0,.9)\n",
    "\n",
    "        hirel= interaction_rel(fg,rels,only_atoms=True)[samples].mean()\n",
    "        if np.isclose(hirel,0,atol=1e-05):\n",
    "            pass\n",
    "        else:\n",
    "            ax.bar(x, \n",
    "                   hirel,\n",
    "                   alpha=0.8, \n",
    "                   color=color,\n",
    "                  align='center',\n",
    "                  width=0.9)\n",
    "\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['bottom'].set_visible(False)\n",
    "    ax.spines['left'].set_visible(False)\n",
    "\n",
    "    ax.set_xlim(-.8,len(subfgs)+.1)\n",
    "\n",
    "    ax.set_xticks(list(xs))\n",
    "    ax.set_xticklabels(names)\n",
    "    ax.set_yticks([])\n",
    "    ax.hlines(0, -.8, len(subfgs), color='black', lw=.8)\n",
    "\n",
    "    color_patches.append(mpatches.Patch(color=(0.,.5,.0,1), label='higher-order'))\n",
    "    color_patches.append(mpatches.Patch(color=(0.9,.65,.0,1), label='lower-order'))\n",
    "\n",
    "\n",
    "    ax.set_ylim(-0.1,5.2)\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(f'imgs/histogram_{order}_mol.png',transparent=True)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### higher and lower order heat maps of mutagenic submolecules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x333.36 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "all_subfgs = []\n",
    "\n",
    "subgraph1 = tuple([0,1,3,5,7,9])\n",
    "subgraph2 = tuple([11,14,15])\n",
    "subgraph3 = tuple([2,4,6,8,10,12,13,16])\n",
    "\n",
    "fgs = subgraph1+ subgraph2\n",
    "\n",
    "order = 1\n",
    "\n",
    "idx = 22\n",
    "\n",
    "model_depth = len(nn.blocks)-1\n",
    "gammas = np.linspace(3,0,model_depth)\n",
    "g = graphs[idx]\n",
    "node_candidates = np.arange(g.nbnodes)\n",
    "shape = (len(node_candidates),)*(model_depth +1)\n",
    "rel_tensor = torch.zeros(*shape)\n",
    "\n",
    "walks = get_all_walks(model_depth+1, g.adj)\n",
    "\n",
    "for walk in walks:\n",
    "    full_relevances = []\n",
    "    rel = nn.lrp(g.get_adj(),gammas,g.label,walk,H0=g.node_features).sum()\n",
    "    full_relevances.append([walk, rel.numpy()])\n",
    "    rel_tensor[tuple(walk)] = rel\n",
    "\n",
    "\n",
    "sub_tensor = rel_tensor.clone()\n",
    "for _ in range((model_depth + 1) - order) : sub_tensor = sub_tensor.sum(-1)\n",
    "\n",
    "sub_relevances = []\n",
    "for walk in get_all_walks(order, g.get_adj()):\n",
    "\n",
    "    index = tuple(set(walk)) if order==1 else tuple(walk)\n",
    "    node_tags = [g.node_tags[node_id] for node_id in walk]\n",
    "    if all([node in fgs for node in walk]):# and len(tuple(set(node_tags))) >1 :\n",
    "        sub_relevances.append([walk, sub_tensor[index].numpy()])\n",
    "\n",
    "plot_mutagenicity(g,sub_relevances,factor=1,figname='imgs/low_order_mutag_mol.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "8806367dad1e0a2cfd7da8c1ed95d51b1c4f8061d16620a9f0a678416f1237e6"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}