Analyze Model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ad14e097",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import os.path as osp\n",
    "\n",
    "import torch\n",
    "from torch.optim import Adam\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.metrics import *\n",
    "\n",
    "import pydgn\n",
    "from pydgn.data.dataset import TUDatasetInterface, DatasetInterface\n",
    "from pydgn.experiment.util import s2c  # string to class\n",
    "\n",
    "import sys\n",
    "\n",
    "import seaborn as sns\n",
    "sns.set_palette(\"colorblind\")\n",
    "\n",
    "plt.rcParams.update({\"text.usetex\": True})\n",
    "plt.rcParams['font.size'] = 12\n",
    "\n",
    "axes_font_size = 18\n",
    "\n",
    "import dgn\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5836c3e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_model(results_folder, model, dataset, dim_node_features, dim_edge_features, num_classes, udn : bool):\n",
    "    if udn:\n",
    "        prefix = 'udn_'\n",
    "    else:\n",
    "        prefix = ''\n",
    "    \n",
    "    best_config_dgn = json.load(open(results_folder + f\"{prefix}{model}_{dataset}/MODEL_ASSESSMENT/OUTER_FOLD_1/MODEL_SELECTION/winner_config.json\",\"r\"))['config']['supervised_config']\n",
    "    final_model_ckpt_dgn = torch.load(results_folder + f\"{prefix}{model}_{dataset}/MODEL_ASSESSMENT/OUTER_FOLD_1/final_run1/best_checkpoint.pth\", map_location=torch.device('cpu'))\n",
    "\n",
    "\n",
    "    model_dgn = s2c(best_config_dgn['model'])(dim_node_features, dim_edge_features, num_classes, None, best_config_dgn)\n",
    "    model_dgn.eval()\n",
    "    model_dgn.to('cpu')\n",
    "    model_dgn.load_state_dict(final_model_ckpt_dgn['model_state'])\n",
    "\n",
    "    return model_dgn"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9aa2697a",
   "metadata": {},
   "source": [
    "## Prepare Data Loaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "683c66d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pydgn.data.provider import DataProvider\n",
    "from torch_geometric.loader import DataLoader\n",
    "\n",
    "def load_loaders(dataset_root, dataset_class, dataset, batch_size=256):\n",
    "\n",
    "    shuffle = False\n",
    "    \n",
    "    \n",
    "    splits_filepath = f'DATA_SPLITS/{dataset}/{dataset}_outer1_inner1.splits'\n",
    "    \n",
    "    dataset_getter = DataProvider(dataset_root, splits_filepath, \n",
    "                                  s2c(dataset_class), dataset,\n",
    "                                  s2c(data_loader_class), {},\n",
    "                                  outer_folds=1, inner_folds=1)\n",
    "    dataset_getter.set_exp_seed(42)\n",
    "    dataset_getter.set_inner_k(0)\n",
    "    dataset_getter.set_outer_k(0)\n",
    "    \n",
    "    # Instantiate the Data Loaders\n",
    "    data_loader_args = dict(batch_size=batch_size, shuffle=shuffle)\n",
    "    \n",
    "    train_loader = dataset_getter.get_outer_train(**data_loader_args)\n",
    "    val_loader = dataset_getter.get_outer_val(**data_loader_args)\n",
    "    test_loader = dataset_getter.get_outer_test(**data_loader_args)\n",
    "\n",
    "    return train_loader, val_loader, test_loader\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa43f13f",
   "metadata": {},
   "source": [
    "## Obtain message filters for model on all nodes and layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dc23bd83",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_filters_per_layer(model, device, train_loader, val_loader, test_loader):\n",
    "    message_filters_all_layers = []\n",
    "\n",
    "    model.eval()\n",
    "    model.to(device)\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        for loader in [train_loader, val_loader, test_loader]:\n",
    "            for data in loader:\n",
    "                data.to(device)\n",
    "        \n",
    "                edge_index = data.edge_index\n",
    "                batch = data.batch\n",
    "                \n",
    "                preds, embeddings, extra = model(data)\n",
    "\n",
    "                if extra[-1] is not None:\n",
    "                    message_filters_all_layers.append(extra[-1].detach().cpu())\n",
    "                else:\n",
    "                    return None\n",
    "        \n",
    "        message_filters = torch.cat(message_filters_all_layers, dim=0)\n",
    "        \n",
    "        return 100*(message_filters.sum((0, 2)) / (message_filters.shape[0]*message_filters.shape[2]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4a297d7-12b1-4d61-a8b7-d3d63006e957",
   "metadata": {},
   "source": [
    "### Percentage of Total Information that is propagated at each layer of AMP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2fb11ca3-9236-4335-91a0-edb86907fab5",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_root = 'DATA'\n",
    "dataset_class = {'Eccentricity': 'dataset.GraphPropertyPrediction',\n",
    "                     'Diameter': 'dataset.GraphPropertyPrediction',\n",
    "                     'SSSP': 'dataset.GraphPropertyPrediction',\n",
    "                     'peptides-func': 'dataset.Peptides',\n",
    "                     'peptides-struct': 'dataset.Peptides'}\n",
    "\n",
    "data_loader_class = 'torch_geometric.loader.DataLoader'\n",
    "device = 'cuda:0'\n",
    "\n",
    "dim_node_features = {'Eccentricity': 2,\n",
    "                     'Diameter': 2,\n",
    "                     'SSSP': 2,\n",
    "                     'peptides-func': 9,\n",
    "                     'peptides-struct': 9}\n",
    "dim_edge_features = 0\n",
    "num_classes = {'Eccentricity': 1,\n",
    "                     'Diameter': 1,\n",
    "                     'SSSP': 1,\n",
    "                     'peptides-func': 10,\n",
    "                     'peptides-struct': 11}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "6d3f568c-ce17-4c5d-99e2-86d2280115a2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing gcn on Eccentricity\n",
      "Processed gcn on Eccentricity\n",
      "Processing gcn on Diameter\n",
      "Processed gcn on Diameter\n",
      "Processing gcn on SSSP\n",
      "Processed gcn on SSSP\n",
      "Processing gcn on peptides-func\n",
      "Processed gcn on peptides-func\n",
      "Processing gcn on peptides-struct\n",
      "Processed gcn on peptides-struct\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "\n",
    "for dataset in ['Eccentricity', 'Diameter', 'SSSP' , 'peptides-func', 'peptides-struct']:\n",
    "\n",
    "    models = ['gcn'] if dataset in ['Eccentricity', 'Diameter', 'SSSP'] else ['gcn']\n",
    "\n",
    "    results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/' if dataset in ['Eccentricity', 'Diameter', 'SSSP'] else '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "    \n",
    "    for model in models:\n",
    "        print(f'Processing {model} on {dataset}')\n",
    "        \n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset)\n",
    "        \n",
    "        message_filters = compute_filters_per_layer(udn_model, device, train_loader, val_loader, test_loader)\n",
    "        \n",
    "        if message_filters is not None:\n",
    "            plt.plot(torch.arange(message_filters.shape[0]) + 1, message_filters,\n",
    "                    label=rf'{dataset}')\n",
    "        else:\n",
    "            print(f'Message filters is none for {dataset}')\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "title = str(r'\\textsc{AMP}$_{\\textsc{%s}}$' % model.upper())\n",
    "plt.title(title, fontsize=axes_font_size, y=1.02)\n",
    "\n",
    "# Set font properties for the axes labels\n",
    "font_properties = {'size': axes_font_size}\n",
    "plt.xlabel('Layers', fontdict=font_properties)\n",
    "plt.ylabel('\\% of information preserved', fontdict=font_properties)\n",
    "\n",
    "# Set font properties for the tick labels\n",
    "plt.xticks(fontsize=axes_font_size)\n",
    "plt.yticks(fontsize=axes_font_size)\n",
    "\n",
    "plt.ylim((0.,100.))\n",
    "\n",
    "plt.legend()\n",
    "plt.tight_layout()\n",
    "plt.savefig(f'images/AMP_{model}_information_preserved.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "abb7ef96-3ef8-4827-bef4-9d685a1cb451",
   "metadata": {},
   "source": [
    "## Layer Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "d22c14be-28ba-4645-879f-f2f0fc7ac80a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing gcn on Eccentricity\n",
      "Processed gcn on Eccentricity\n",
      "Processing gin on Eccentricity\n",
      "Processed gin on Eccentricity\n",
      "Processing adgn on Eccentricity\n",
      "Processed adgn on Eccentricity\n",
      "Processing gcn on Diameter\n",
      "Processed gcn on Diameter\n",
      "Processing gin on Diameter\n",
      "Processed gin on Diameter\n",
      "Processing adgn on Diameter\n",
      "Processed adgn on Diameter\n",
      "Processing gcn on SSSP\n",
      "Processed gcn on SSSP\n",
      "Processing gin on SSSP\n",
      "Processed gin on SSSP\n",
      "Processing adgn on SSSP\n",
      "Processed adgn on SSSP\n",
      "Processing gcn on peptides-func\n",
      "Processed gcn on peptides-func\n",
      "Processing gine on peptides-func\n",
      "Processed gine on peptides-func\n",
      "Processing gatedgcn on peptides-func\n",
      "Processed gatedgcn on peptides-func\n",
      "Processing gcn on peptides-struct\n",
      "Processed gcn on peptides-struct\n",
      "Processing gine on peptides-struct\n",
      "Processed gine on peptides-struct\n",
      "Processing gatedgcn on peptides-struct\n",
      "Processed gatedgcn on peptides-struct\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x494 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6.4, 4.94))\n",
    "\n",
    "# Set the width of the bars\n",
    "bar_width = 0.2\n",
    "\n",
    "c = 0\n",
    "for dataset in ['Eccentricity', 'Diameter', 'SSSP']:\n",
    "\n",
    "    models = ['gcn', 'gin', 'adgn'] \n",
    "\n",
    "    for model in models:\n",
    "\n",
    "        results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/' if model != 'gin' else '/mnt/local/ferrica/UDN_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "        \n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        \n",
    "        qL_probs = udn_model.variational_L.compute_probability_vector().detach().numpy()\n",
    "\n",
    "        if dataset == 'Eccentricity':\n",
    "            dataset_str = 'E'\n",
    "        elif dataset == 'Diameter':\n",
    "            dataset_str = 'D'\n",
    "        elif dataset == 'SSSP':\n",
    "            dataset_str = 'S'\n",
    "        plt.bar(torch.arange(qL_probs.shape[0]) + 1 + bar_width*c, qL_probs, \n",
    "                width=bar_width, \n",
    "                label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset_str, model.upper()), \n",
    "                alpha=0.5)\n",
    "\n",
    "        c += 1\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "\n",
    "plt.title('Eccentr. (E), Diameter (D), SSSP (S)', fontsize=axes_font_size, y=1.02)\n",
    "\n",
    "# Set font properties for the axes labels\n",
    "font_properties = {'size': axes_font_size}\n",
    "plt.xlabel('Layers', fontdict=font_properties)\n",
    "plt.ylabel('Probability', fontdict=font_properties)\n",
    "\n",
    "# Set font properties for the tick labels\n",
    "plt.xticks(fontsize=axes_font_size)\n",
    "plt.yticks(fontsize=axes_font_size)\n",
    "\n",
    "legend = plt.legend(loc='upper center', bbox_to_anchor=(0.5, -0.2), ncol=3, columnspacing=0.5)\n",
    "# plt.tight_layout()\n",
    "plt.savefig(f'images/AMP_distributions_synthetic.pdf', bbox_extra_artists=(legend,), bbox_inches='tight')\n",
    "\n",
    "plt.figure(figsize=(6.4, 4.8))\n",
    "\n",
    "c = 0\n",
    "\n",
    "# Set the width of the bars\n",
    "bar_width = 0.35\n",
    "\n",
    "for dataset in ['peptides-func', 'peptides-struct']:\n",
    "\n",
    "    models = ['gcn', 'gine', 'gatedgcn']\n",
    "\n",
    "    results_folder = '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "    \n",
    "    for model in models:\n",
    "        print(f'Processing {model} on {dataset}')\n",
    "        \n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        \n",
    "        qL_probs = udn_model.variational_L.compute_probability_vector().detach().numpy()\n",
    "\n",
    "\n",
    "        dataset_str = 'F' if (dataset == 'peptides-func') else 'S'\n",
    "        plt.bar(torch.arange(qL_probs.shape[0]) + 1 + bar_width*c, qL_probs, \n",
    "                width=bar_width, \n",
    "                label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset_str, model.upper()), \n",
    "                alpha=0.5)\n",
    "\n",
    "        c += 1\n",
    "        \n",
    "        print(f'Processed {model} on {dataset}')\n",
    "plt.title('peptides func (F) / struct (S)', fontsize=axes_font_size, y=1.02)\n",
    "\n",
    "# Set font properties for the axes labels\n",
    "font_properties = {'size': axes_font_size}\n",
    "plt.xlabel('Layers', fontdict=font_properties)\n",
    "plt.ylabel('Probability', fontdict=font_properties)\n",
    "\n",
    "# Set font properties for the tick labels\n",
    "plt.xticks(fontsize=axes_font_size)\n",
    "plt.yticks(fontsize=axes_font_size)\n",
    "\n",
    "legend = plt.legend(loc='upper center', bbox_to_anchor=(0.5, -0.2), ncol=2, columnspacing=0.5)\n",
    "# plt.tight_layout()\n",
    "plt.savefig(f'images/AMP_distributions_peptides.pdf', bbox_extra_artists=(legend,), bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c6ff8c1-f7bb-41c3-b94e-379f2b0abc4f",
   "metadata": {},
   "source": [
    "## Oversmoothing Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "56f8e2f9-4c82-4236-b931-4a05a7efaba1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_oversmoothing(model, train_loader, val_loader, test_loader):\n",
    "    total_layerwise_energies = 0.\n",
    "    num_total_nodes = 0.\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        for loader in [train_loader, val_loader, test_loader]:\n",
    "            for data in loader:\n",
    "                data.to(device)\n",
    "        \n",
    "                edge_index = data.edge_index\n",
    "                batch = data.batch\n",
    "                _, embeddings, _ = model(data)\n",
    "                \n",
    "                sq_node_diff = torch.sum((embeddings[edge_index[0]] - embeddings[edge_index[1]])**2, dim=-1)\n",
    "                total_layerwise_energies += sq_node_diff.sum(0)\n",
    "                num_total_nodes += sq_node_diff.shape[0] \n",
    "                \n",
    "    layerwise_energies = torch.sqrt(total_layerwise_energies / num_total_nodes)\n",
    "    return layerwise_energies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "16158ec9-e92b-4fdc-a8cc-a42163e010a3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing gcn on Eccentricity\n"
     ]
    },
    {
     "ename": "NameError",
     "evalue": "name 'compute_oversmoothing' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001B[0;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[0;31mNameError\u001B[0m                                 Traceback (most recent call last)",
      "Cell \u001B[0;32mIn[35], line 20\u001B[0m\n\u001B[1;32m     16\u001B[0m udn_model\u001B[38;5;241m.\u001B[39mto(device)\n\u001B[1;32m     18\u001B[0m train_loader, val_loader, test_loader \u001B[38;5;241m=\u001B[39m load_loaders(dataset_root, dataset_class[dataset], dataset)\n\u001B[0;32m---> 20\u001B[0m layerwise_energies \u001B[38;5;241m=\u001B[39m \u001B[43mcompute_oversmoothing\u001B[49m(udn_model, train_loader, val_loader, test_loader)\u001B[38;5;241m.\u001B[39mcpu()\u001B[38;5;241m.\u001B[39mdetach()\u001B[38;5;241m.\u001B[39mnumpy()\n\u001B[1;32m     21\u001B[0m base_layerwise_energies \u001B[38;5;241m=\u001B[39m compute_oversmoothing(base_model, train_loader, val_loader, test_loader)\u001B[38;5;241m.\u001B[39mcpu()\u001B[38;5;241m.\u001B[39mdetach()\u001B[38;5;241m.\u001B[39mnumpy()\n\u001B[1;32m     23\u001B[0m plt\u001B[38;5;241m.\u001B[39mplot(torch\u001B[38;5;241m.\u001B[39marange(base_layerwise_energies\u001B[38;5;241m.\u001B[39mshape[\u001B[38;5;241m0\u001B[39m]) \u001B[38;5;241m+\u001B[39m \u001B[38;5;241m1\u001B[39m, np\u001B[38;5;241m.\u001B[39mlog(base_layerwise_energies), label\u001B[38;5;241m=\u001B[39m\u001B[38;5;124mr\u001B[39m\u001B[38;5;124m'\u001B[39m\u001B[38;5;132;01m%s\u001B[39;00m\u001B[38;5;124m (\u001B[39m\u001B[38;5;124m\\\u001B[39m\u001B[38;5;124mtextsc\u001B[39m\u001B[38;5;124m{\u001B[39m\u001B[38;5;132;01m%s\u001B[39;00m\u001B[38;5;124m})\u001B[39m\u001B[38;5;124m'\u001B[39m \u001B[38;5;241m%\u001B[39m (dataset, model\u001B[38;5;241m.\u001B[39mupper()), linestyle\u001B[38;5;241m=\u001B[39m\u001B[38;5;124m'\u001B[39m\u001B[38;5;124m--\u001B[39m\u001B[38;5;124m'\u001B[39m)\n",
      "\u001B[0;31mNameError\u001B[0m: name 'compute_oversmoothing' is not defined"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "\n",
    "for dataset in ['Eccentricity', 'Diameter', 'SSSP']:\n",
    "\n",
    "    models = ['gcn']#, 'gin', 'adgn'] \n",
    "\n",
    "    for model in models:\n",
    "\n",
    "        results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/' if model != 'gin' else '/mnt/local/ferrica/UDN_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "\n",
    "        base_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=False)\n",
    "        base_model.to(device)\n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        udn_model.to(device)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset)\n",
    "\n",
    "        layerwise_energies = compute_oversmoothing(udn_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        base_layerwise_energies = compute_oversmoothing(base_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        \n",
    "        plt.plot(torch.arange(base_layerwise_energies.shape[0]) + 1, np.log(base_layerwise_energies), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        plt.plot(torch.arange(layerwise_energies.shape[0]) + 1, np.log(layerwise_energies), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "\n",
    "\n",
    "for dataset in ['peptides-func', 'peptides-struct']:\n",
    "\n",
    "    models = ['gcn']#, 'gin', 'adgn'] \n",
    "\n",
    "    for model in models:\n",
    "\n",
    "        results_folder = '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "\n",
    "        base_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=False)\n",
    "        base_model.to(device)\n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        udn_model.to(device)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset)\n",
    "\n",
    "        layerwise_energies = compute_oversmoothing(udn_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        base_layerwise_energies = compute_oversmoothing(base_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        \n",
    "        plt.plot(torch.arange(base_layerwise_energies.shape[0]) + 1, np.log(base_layerwise_energies), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        plt.plot(torch.arange(layerwise_energies.shape[0]) + 1, np.log(layerwise_energies), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "\n",
    "# Set font properties for the axes labels\n",
    "font_properties = {'size': axes_font_size}\n",
    "plt.xlabel('Layers', fontdict=font_properties)\n",
    "plt.ylabel(r'$\\ln$ Dirichlet Energy', fontdict=font_properties)\n",
    "\n",
    "# Set font properties for the tick labels\n",
    "plt.xticks(fontsize=axes_font_size)\n",
    "plt.yticks(fontsize=axes_font_size)\n",
    "\n",
    "plt.legend(bbox_to_anchor=(1.04, 1), loc=\"upper left\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(f'images/Dirichlet_Energy.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f41d14d5-25da-4d4e-a19c-4760effec1d6",
   "metadata": {},
   "source": [
    "## Oversquashing Analysis (slow, only on validation loader!)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "78fdaf9c-3f6f-49f9-967c-ad2db2c82ff8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_oversquashing(model, loader, total_nodes_to_process=10):\n",
    "\n",
    "    sensitivity_matrix = None\n",
    "    print(f'Need to process {len(loader)} batches...')\n",
    "    c = 0\n",
    "    for data in loader:\n",
    "        data.to(device)\n",
    "        data.x.requires_grad=True\n",
    "        batch = data.batch\n",
    "        max_batch = batch.max()\n",
    "        num_nodes, _ = data.x.shape\n",
    "\n",
    "        preds, embeddings, extra = model(data, retain_grad=True)\n",
    "\n",
    "        if sensitivity_matrix is None:\n",
    "            sensitivity_matrix = torch.zeros(len(embeddings))\n",
    "        \n",
    "        for nodeid in range(num_nodes):\n",
    "            print(nodeid)\n",
    "            # print(preds.shape)\n",
    "            for f in range(embeddings[-1].shape[1]):\n",
    "                embeddings[-1][nodeid, f].backward(retain_graph=True)\n",
    "        \n",
    "                # layer_wise_sensitivity = []\n",
    "                for hidden_layer in range(len(embeddings)):\n",
    "                    if embeddings[hidden_layer].grad != None:\n",
    "                        sensitivity_matrix[hidden_layer] += torch.norm(embeddings[hidden_layer].grad, p=1).cpu()\n",
    "                        embeddings[hidden_layer].grad = None\n",
    "\n",
    "            c += 1\n",
    "            if c == total_nodes_to_process:\n",
    "                break\n",
    "\n",
    "        if c == total_nodes_to_process:\n",
    "            print(f'Processed {c} nodes')\n",
    "            print('Stopping here for computational reasons!')\n",
    "            break\n",
    "\n",
    "        break\n",
    "    return sensitivity_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "23cb9fae-8a82-4cbb-8910-b475c0ccaa73",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing gcn on Eccentricity\n",
      "Need to process 20 batches...\n",
      "0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_434530/2628889754.py:26: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:486.)\n",
      "  if embeddings[hidden_layer].grad != None:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Need to process 20 batches...\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Processed gcn on Eccentricity\n",
      "Processing gcn on Diameter\n",
      "Need to process 20 batches...\n",
      "0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_434530/2628889754.py:26: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:486.)\n",
      "  if embeddings[hidden_layer].grad != None:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Need to process 20 batches...\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Processed gcn on Diameter\n",
      "Processing gcn on SSSP\n",
      "Need to process 20 batches...\n",
      "0\n",
      "1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_434530/2628889754.py:26: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:486.)\n",
      "  if embeddings[hidden_layer].grad != None:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Need to process 20 batches...\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Processed gcn on SSSP\n",
      "Processing gcn on peptides-func\n",
      "Need to process 146 batches...\n",
      "0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_434530/2628889754.py:26: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:486.)\n",
      "  if embeddings[hidden_layer].grad != None:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Need to process 146 batches...\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Processed gcn on peptides-func\n",
      "Processing gcn on peptides-struct\n",
      "Need to process 146 batches...\n",
      "0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_434530/2628889754.py:26: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:486.)\n",
      "  if embeddings[hidden_layer].grad != None:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Need to process 146 batches...\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n",
      "50\n",
      "51\n",
      "52\n",
      "53\n",
      "54\n",
      "55\n",
      "56\n",
      "57\n",
      "58\n",
      "59\n",
      "60\n",
      "61\n",
      "62\n",
      "63\n",
      "64\n",
      "65\n",
      "66\n",
      "67\n",
      "68\n",
      "69\n",
      "70\n",
      "71\n",
      "72\n",
      "73\n",
      "74\n",
      "75\n",
      "76\n",
      "77\n",
      "78\n",
      "79\n",
      "80\n",
      "81\n",
      "82\n",
      "83\n",
      "84\n",
      "85\n",
      "86\n",
      "87\n",
      "88\n",
      "89\n",
      "90\n",
      "91\n",
      "92\n",
      "93\n",
      "94\n",
      "95\n",
      "96\n",
      "97\n",
      "98\n",
      "99\n",
      "100\n",
      "101\n",
      "102\n",
      "103\n",
      "104\n",
      "105\n",
      "106\n",
      "107\n",
      "108\n",
      "109\n",
      "110\n",
      "111\n",
      "112\n",
      "113\n",
      "114\n",
      "115\n",
      "116\n",
      "117\n",
      "118\n",
      "119\n",
      "120\n",
      "121\n",
      "122\n",
      "123\n",
      "124\n",
      "125\n",
      "126\n",
      "127\n",
      "128\n",
      "129\n",
      "130\n",
      "131\n",
      "132\n",
      "133\n",
      "134\n",
      "135\n",
      "136\n",
      "137\n",
      "138\n",
      "139\n",
      "140\n",
      "141\n",
      "142\n",
      "143\n",
      "144\n",
      "145\n",
      "146\n",
      "147\n",
      "148\n",
      "149\n",
      "150\n",
      "151\n",
      "152\n",
      "153\n",
      "154\n",
      "155\n",
      "156\n",
      "157\n",
      "158\n",
      "159\n",
      "160\n",
      "161\n",
      "162\n",
      "163\n",
      "164\n",
      "165\n",
      "166\n",
      "167\n",
      "168\n",
      "169\n",
      "170\n",
      "171\n",
      "172\n",
      "173\n",
      "174\n",
      "175\n",
      "176\n",
      "177\n",
      "178\n",
      "179\n",
      "180\n",
      "181\n",
      "182\n",
      "183\n",
      "184\n",
      "185\n",
      "186\n",
      "187\n",
      "188\n",
      "189\n",
      "190\n",
      "191\n",
      "192\n",
      "193\n",
      "194\n",
      "195\n",
      "196\n",
      "197\n",
      "198\n",
      "199\n",
      "200\n",
      "201\n",
      "202\n",
      "203\n",
      "204\n",
      "205\n",
      "206\n",
      "207\n",
      "208\n",
      "209\n",
      "210\n",
      "211\n",
      "212\n",
      "213\n",
      "214\n",
      "215\n",
      "216\n",
      "217\n",
      "218\n",
      "219\n",
      "220\n",
      "221\n",
      "222\n",
      "223\n",
      "224\n",
      "225\n",
      "226\n",
      "227\n",
      "228\n",
      "229\n",
      "230\n",
      "231\n",
      "232\n",
      "233\n",
      "234\n",
      "235\n",
      "236\n",
      "237\n",
      "238\n",
      "239\n",
      "240\n",
      "241\n",
      "242\n",
      "243\n",
      "244\n",
      "245\n",
      "246\n",
      "247\n",
      "248\n",
      "249\n",
      "Processed 250 nodes\n",
      "Stopping here for computational reasons!\n",
      "Processed gcn on peptides-struct\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "device = 'cuda:0'\n",
    "\n",
    "\n",
    "models = ['gcn'] \n",
    "\n",
    "for model in models:\n",
    "    plt.figure()\n",
    "    for dataset in ['Eccentricity', 'Diameter', 'SSSP', 'peptides-func', 'peptides-struct']:\n",
    "\n",
    "        if dataset in ['Eccentricity', 'Diameter', 'SSSP']:\n",
    "            results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/'\n",
    "        else:\n",
    "            results_folder = '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "\n",
    "        # for some reason one specific GIN checkpoint is not loading\n",
    "        if model == 'gin':\n",
    "            if dataset == 'Eccentricity':\n",
    "                results_folder = '/mnt/local/ferrica/UDN_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "    \n",
    "        base_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=False)\n",
    "        base_model.to(device)\n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        udn_model.to(device)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset, batch_size=32)\n",
    "    \n",
    "        layerwise_sensitivity = compute_oversquashing(udn_model, val_loader, total_nodes_to_process=250).cpu().detach().numpy()\n",
    "        base_layerwise_sensitivity = compute_oversquashing(base_model, val_loader, total_nodes_to_process=250).cpu().detach().numpy()\n",
    "\n",
    "        layerwise_sensitivity[-1] = 1.\n",
    "        base_layerwise_sensitivity[-1] = 1.\n",
    "\n",
    "        torch.save(layerwise_sensitivity, f'images/AMP_{model}_{dataset}_sensitivity.pt')\n",
    "        torch.save(base_layerwise_sensitivity, f'images/base_{model}_{dataset}_sensitivity.pt')\n",
    "        \n",
    "        plt.plot(torch.arange(base_layerwise_sensitivity.shape[0]-1) + 1, np.log(base_layerwise_sensitivity[:-1]), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        plt.plot(torch.arange(layerwise_sensitivity.shape[0]-1) + 1, np.log(layerwise_sensitivity[:-1]), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "        \n",
    "        print(f'Processed {model} on {dataset}')\n",
    "        \n",
    "    # Set font properties for the axes labels\n",
    "    font_properties = {'size': axes_font_size}\n",
    "    plt.xlabel('Layers', fontdict=font_properties)\n",
    "    plt.ylabel(r'$\\ln$ Layer-wise Sensitivity', fontdict=font_properties)\n",
    "    \n",
    "    # Set font properties for the tick labels\n",
    "    plt.xticks(fontsize=axes_font_size)\n",
    "    plt.yticks(fontsize=axes_font_size)\n",
    "    \n",
    "    plt.legend(bbox_to_anchor=(1.04, 1), loc=\"upper left\")    \n",
    "    plt.tight_layout()\n",
    "    plt.savefig(f'images/AMP_{model}_sensitivity.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "af4ad3ca-52e8-4ef9-9790-63ee0b57fce9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing gcn on Eccentricity\n",
      "Processed gcn on Eccentricity\n",
      "Processing gcn on Diameter\n",
      "Processed gcn on Diameter\n",
      "Processing gcn on SSSP\n",
      "Processed gcn on SSSP\n",
      "Processing gcn on peptides-func\n",
      "Processed gcn on peptides-func\n",
      "Processing gcn on peptides-struct\n",
      "Processed gcn on peptides-struct\n",
      "Processing gcn on Eccentricity\n",
      "Processed gcn on Eccentricity\n",
      "Processing gcn on Diameter\n",
      "Processed gcn on Diameter\n",
      "Processing gcn on SSSP\n",
      "Processed gcn on SSSP\n",
      "Processing gcn on peptides-func\n",
      "Processed gcn on peptides-func\n",
      "Processing gcn on peptides-struct\n",
      "Processed gcn on peptides-struct\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1050x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols=2, figsize=(10.5,4.8))\n",
    "\n",
    "for dataset in ['Eccentricity', 'Diameter', 'SSSP']:\n",
    "\n",
    "    models = ['gcn']#, 'gin', 'adgn'] \n",
    "\n",
    "    for model in models:\n",
    "\n",
    "        results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/' if model != 'gin' else '/mnt/local/ferrica/UDN_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "\n",
    "        base_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=False)\n",
    "        base_model.to(device)\n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        udn_model.to(device)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset)\n",
    "\n",
    "        layerwise_energies = compute_oversmoothing(udn_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        base_layerwise_energies = compute_oversmoothing(base_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        \n",
    "        axes[0].plot(torch.arange(base_layerwise_energies.shape[0]) + 1, np.log(base_layerwise_energies), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        axes[0].plot(torch.arange(layerwise_energies.shape[0]) + 1, np.log(layerwise_energies), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "\n",
    "\n",
    "for dataset in ['peptides-func', 'peptides-struct']:\n",
    "\n",
    "    models = ['gcn']#, 'gin', 'adgn'] \n",
    "\n",
    "    for model in models:\n",
    "\n",
    "        results_folder = '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "\n",
    "        base_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=False)\n",
    "        base_model.to(device)\n",
    "        udn_model = load_model(results_folder, model, dataset, dim_node_features[dataset], dim_edge_features, num_classes[dataset], udn=True)\n",
    "        udn_model.to(device)\n",
    "        \n",
    "        train_loader, val_loader, test_loader = load_loaders(dataset_root, dataset_class[dataset], dataset)\n",
    "\n",
    "        layerwise_energies = compute_oversmoothing(udn_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        base_layerwise_energies = compute_oversmoothing(base_model, train_loader, val_loader, test_loader).cpu().detach().numpy()\n",
    "        \n",
    "        axes[0].plot(torch.arange(base_layerwise_energies.shape[0]) + 1, np.log(base_layerwise_energies), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        axes[0].plot(torch.arange(layerwise_energies.shape[0]) + 1, np.log(layerwise_energies), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "\n",
    "        # Set font properties for the tick labels\n",
    "        plt.xticks(fontsize=axes_font_size)\n",
    "        plt.yticks(fontsize=axes_font_size)\n",
    "\n",
    "        axes[0].tick_params(axis='both', which='major', labelsize=axes_font_size)\n",
    "\n",
    "        # Set font properties for the axes labels\n",
    "        font_properties = {'size': axes_font_size}\n",
    "        axes[0].set_xlabel('Layers', fontdict=font_properties)\n",
    "        axes[0].set_ylabel(r'$\\ln$ Layer-wise Dirichlet Energy', fontdict=font_properties)\n",
    "\n",
    "        print(f'Processed {model} on {dataset}')\n",
    "\n",
    "models = ['gcn'] \n",
    "\n",
    "for model in models:\n",
    "    for dataset in ['Eccentricity', 'Diameter', 'SSSP', 'peptides-func', 'peptides-struct']:\n",
    "\n",
    "        if dataset in ['Eccentricity', 'Diameter', 'SSSP']:\n",
    "            results_folder = '/mnt/local/ferrica/UDN_RESULTS_300PATIENCE/'\n",
    "        else:\n",
    "            results_folder = '/mnt/local/ferrica/LRGB_RESULTS/'\n",
    "\n",
    "        # for some reason one specific GIN checkpoint is not loading\n",
    "        if model == 'gin':\n",
    "            if dataset == 'Eccentricity':\n",
    "                results_folder = '/mnt/local/ferrica/UDN_RESULTS/'\n",
    "        \n",
    "        print(f'Processing {model} on {dataset}')\n",
    "\n",
    "        with torch.no_grad():\n",
    "            layerwise_sensitivity = torch.load(f'images/AMP_{model}_{dataset}_sensitivity.pt')\n",
    "            base_layerwise_sensitivity = torch.load(f'images/base_{model}_{dataset}_sensitivity.pt')\n",
    "\n",
    "        \n",
    "        axes[1].plot(torch.arange(base_layerwise_sensitivity.shape[0]-1) + 1, np.log(base_layerwise_sensitivity[:-1]), label=r'%s (\\textsc{%s})' % (dataset, model.upper()), linestyle='--')\n",
    "        axes[1].plot(torch.arange(layerwise_sensitivity.shape[0]-1) + 1, np.log(layerwise_sensitivity[:-1]), label=r'%s (\\textsc{AMP}$_{\\textsc{%s}}$)' % (dataset, model.upper()))\n",
    "        \n",
    "        print(f'Processed {model} on {dataset}')\n",
    "        \n",
    "    # Set font properties for the axes labels\n",
    "    font_properties = {'size': axes_font_size}\n",
    "    plt.xlabel('Layers', fontdict=font_properties)\n",
    "    plt.ylabel(r'$\\ln$ Layer-wise Sensitivity', fontdict=font_properties)\n",
    "    \n",
    "    # Set font properties for the tick labels\n",
    "    plt.xticks(fontsize=axes_font_size)\n",
    "    plt.yticks(fontsize=axes_font_size)  \n",
    "\n",
    "legend = plt.legend(bbox_to_anchor=(1.04, 1), loc=\"upper left\")\n",
    "# plt.tight_layout()\n",
    "plt.savefig(f'images/oversmoothing_oversquashing.pdf', bbox_extra_artists=(legend,), bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2d36b8c0-bdf1-4ffb-a2e6-4d0d3c3be94c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}