base-surrogate-steam.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Library imports**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Library imports\n",
    "# import pyforest\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "from matplotlib import pyplot as plt\n",
    "from tqdm import tqdm\n",
    "from pprint import pprint\n",
    "from time import sleep\n",
    "import time\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "from turtle import forward\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import math\n",
    "import torch\n",
    "from torch.nn.parameter import Parameter\n",
    "from torch.nn.modules.module import Module\n",
    "from torch.utils.data import DataLoader, TensorDataset\n",
    "\n",
    "import models \n",
    "from models import get_accuracy\n",
    "import inspect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Load dataset**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1698, 1397, 168848, 168848)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = 'steam'\n",
    "\n",
    "train_edges = np.load('data/' + dataset + '/train_edges.npy')\n",
    "test_edges = np.load('data/' + dataset + '/train_edges.npy')\n",
    "\n",
    "user_list_train = train_edges[:, 0]\n",
    "user_list_test = test_edges[:, 0]\n",
    "item_list_train = train_edges[:, 1]\n",
    "item_list_test = test_edges[:, 1]\n",
    "rating_list_train = train_edges[:, 2].astype('float32')\n",
    "rating_list_test = test_edges[:, 2].astype('float32')\n",
    "\n",
    "n_users = max(user_list_train.max(), user_list_test.max()) + 1 \n",
    "n_items = max(item_list_train.max(), item_list_test.max()) + 1\n",
    "n_samples_train = len(rating_list_train)\n",
    "n_samples_test = len(rating_list_test)\n",
    "\n",
    "n_users, n_items, n_samples_train, n_samples_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Check available models**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['CF', 'CFD', 'GMF', 'GMFB', 'GMFD', 'MLP', 'NeuMF']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[m[0] for m in inspect.getmembers(models, inspect.isclass) if m[1].__module__ == 'models']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Code for surrogate meta-attack (surrogate-CF-SGD, evaluation-CF-Adam, same-init)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> Algorithm:  base\n",
      "\n",
      "-> T (surrogate):  50\n",
      "-> T (target):  50\n",
      "-> Delta: 40000 (23.69%)\n",
      "-> Embedding size (surrogate):  32\n",
      "-> Embedding size (target):  32\n",
      "-> Device:  cuda:2\n",
      "\n",
      "-> Surrogate:  CFD\n",
      "-> Target:  CFD\n",
      "-> Surrogate optimizer:  adam\n",
      "-> Target optimizer:  adam\n",
      "-> Surrogate learning rate:  0.3\n",
      "-> Target learning rate:  0.3\n",
      "-> Surrogate seed:  0\n",
      "-> Target seed:  1\n",
      "\n",
      "-> Retain graph:  True\n",
      "-> Create graph:  False\n",
      "-> Save results:  True\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "-> Perturbations: 100%|██████████| 40000/40000 [6:21:13<00:00,  1.75it/s]  \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> Execution time: 06h 21m 14s\n",
      "-> File name: extra-steam-CFD(32,adam)-CFD(32,adam)-base-D=40000-T(50,50)-lr(0.3,0.3)-seed(0,1)-drop(0.2,0.2)\n"
     ]
    }
   ],
   "source": [
    "# model settings\n",
    "algorithm = 'base'\n",
    "surrogate = 'CFD'\n",
    "target = 'CFD'\n",
    "opt_surrogate = 'adam'\n",
    "opt_target = 'adam'\n",
    "lr_surrogate = 0.3\n",
    "lr_target = 0.3\n",
    "seed_surrogate = 0\n",
    "seed_target = 1\n",
    "dropout_surrogate = 0.2\n",
    "dropout_target = 0.2\n",
    "\n",
    "# start execution\n",
    "start_time = time.time()\n",
    "\n",
    "# GPU settings (set use_gpu = -1 if you want to use CPU)\n",
    "use_gpu = 2\n",
    "if use_gpu == -1:\n",
    "    device = 'cpu'\n",
    "else:\n",
    "    device = torch.device('cuda:{}'.format(str(use_gpu)) if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "# some hyperparameters\n",
    "T_surrogate = 50\n",
    "T_target = 50\n",
    "Delta = 40000 # 5% ~ 10K perturbations for movielens\n",
    "n_factors_surrogate = 32\n",
    "n_factors_target = 32\n",
    "save_results = True\n",
    "retain_graph = True \n",
    "create_graph = False\n",
    "\n",
    "if Delta < 100:\n",
    "    save_results = False\n",
    "\n",
    "# initialize list of perturbations\n",
    "perturbations = dict()\n",
    "perturbations['edges'] = []\n",
    "perturbations['metagrad'] = []\n",
    "\n",
    "perturbations['accuracy-before-surrogate'] = []\n",
    "perturbations['accuracy-after-surrogate'] = []\n",
    "perturbations['accuracy-unseen-surrogate'] = []\n",
    "\n",
    "perturbations['loss-before-surrogate'] = []\n",
    "perturbations['loss-after-surrogate'] = []\n",
    "perturbations['loss-unseen-surrogate'] = []\n",
    "\n",
    "perturbations['accuracy-before-target'] = []\n",
    "perturbations['accuracy-after-target'] = []\n",
    "perturbations['accuracy-unseen-target'] = []\n",
    "\n",
    "perturbations['loss-before-target'] = []\n",
    "perturbations['loss-after-target'] = []\n",
    "perturbations['loss-unseen-target'] = []\n",
    "\n",
    "perturbations['auc-before-surrogate'] = []\n",
    "perturbations['auc-after-surrogate'] = []\n",
    "perturbations['auc-unseen-surrogate'] = []\n",
    "\n",
    "perturbations['auc-before-target'] = []\n",
    "perturbations['auc-after-target'] = []\n",
    "perturbations['auc-unseen-target'] = []\n",
    "\n",
    "# print hyperparam configuration\n",
    "print('-> Algorithm: ', algorithm)\n",
    "print()\n",
    "print('-> T (surrogate): ', T_surrogate)\n",
    "print('-> T (target): ', T_target)\n",
    "print('-> Delta: {} ({}%)'.format(Delta, round(Delta * 100 / n_samples_train, 2)))\n",
    "print('-> Embedding size (surrogate): ', n_factors_surrogate)\n",
    "print('-> Embedding size (target): ', n_factors_target)\n",
    "print('-> Device: ', device)\n",
    "print()\n",
    "print('-> Surrogate: ', surrogate)\n",
    "print('-> Target: ', target)\n",
    "print('-> Surrogate optimizer: ', opt_surrogate)\n",
    "print('-> Target optimizer: ', opt_target)\n",
    "print('-> Surrogate learning rate: ', lr_surrogate)\n",
    "print('-> Target learning rate: ', lr_target)\n",
    "print('-> Surrogate seed: ', seed_surrogate)\n",
    "print('-> Target seed: ', seed_target)\n",
    "print()\n",
    "print('-> Retain graph: ', retain_graph)\n",
    "print('-> Create graph: ', create_graph)\n",
    "print('-> Save results: ', save_results)\n",
    "\n",
    "# load users, items and ratings as tensors\n",
    "users = torch.tensor(user_list_train, device = device)\n",
    "items = torch.tensor(item_list_train, device = device)\n",
    "ratings = torch.tensor(rating_list_train, device = device, requires_grad = True)\n",
    "perturbs = torch.ones_like(ratings).bool()\n",
    "\n",
    "users_test = torch.tensor(user_list_test, device = device)\n",
    "items_test = torch.tensor(item_list_test, device = device)\n",
    "ratings_test = torch.tensor(rating_list_test, device = device)\n",
    "\n",
    "# sample random negative edges to perturb\n",
    "if 'base' in algorithm:\n",
    "    edges = ratings.detach().to('cpu').numpy()\n",
    "    neg_edges = np.where(edges == 0)[0]\n",
    "    np.random.seed(0)\n",
    "    edges_to_perturb = np.random.choice(neg_edges, size=Delta, replace = True) # sample Delta edges randomly and perturb one by one inside loop \n",
    "\n",
    "# for each perturbation do the following\n",
    "for delta in tqdm(range(Delta), desc='-> Perturbations'):\n",
    "\n",
    "    # define surrogate model and it's parameters\n",
    "    torch.manual_seed(seed_surrogate)\n",
    "    model = getattr(models, surrogate)(n_users, n_items, n_factors_surrogate, dropout_surrogate).to(device)\n",
    "\n",
    "    # define optimizer and loss function\n",
    "    if 'adam' in opt_surrogate:\n",
    "        optimizer = torch.optim.Adam(model.parameters(), lr = lr_surrogate)\n",
    "    else:\n",
    "        optimizer = torch.optim.SGD(model.parameters(), lr = lr_surrogate)\n",
    "    loss_fn = nn.BCELoss(reduction = 'mean')\n",
    "\n",
    "    # inner loop training process\n",
    "    model.train()\n",
    "    for i in range(T_surrogate):\n",
    "        y_hat = model(users, items).reshape(ratings.shape)\n",
    "        loss = loss_fn(y_hat, ratings)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward(retain_graph=retain_graph, create_graph=create_graph)\n",
    "        optimizer.step()\n",
    "    model.eval()\n",
    "\n",
    "    # compute and store accuracy of model after T training steps\n",
    "    with torch.no_grad():\n",
    "        # training accuracy and loss including perturbed edges\n",
    "        y_hat = model(users, items).reshape(ratings.shape)\n",
    "        perturbations['accuracy-before-surrogate'].append(get_accuracy(y_hat, ratings))\n",
    "        perturbations['loss-before-surrogate'].append(loss_fn(y_hat, ratings).item())\n",
    "\n",
    "        # compute training AUROC including perturbed edges\n",
    "        y_pred = y_hat.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings.detach().clone().to('cpu').numpy() \n",
    "        perturbations['auc-before-surrogate'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "        # training accuracy and loss excluding perturbed edges\n",
    "        y_hat_masked = torch.masked_select(y_hat, perturbs)\n",
    "        ratings_masked = torch.masked_select(ratings, perturbs)\n",
    "        perturbations['accuracy-after-surrogate'].append(get_accuracy(y_hat_masked, ratings_masked))\n",
    "        perturbations['loss-after-surrogate'].append(loss_fn(y_hat_masked, ratings_masked).item())\n",
    "\n",
    "        # compute training AUROC excluding perturbed edges\n",
    "        y_pred = y_hat_masked.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings_masked.detach().clone().to('cpu').numpy()\n",
    "        perturbations['auc-after-surrogate'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "    # compute and store accuracy of surrogate model on unseen data\n",
    "    with torch.no_grad():\n",
    "        y_hat = model(users_test, items_test).reshape(ratings_test.shape)\n",
    "        perturbations['accuracy-unseen-surrogate'].append(get_accuracy(y_hat, ratings_test))\n",
    "        perturbations['loss-unseen-surrogate'].append(loss_fn(y_hat, ratings_test).item())\n",
    "\n",
    "        # compute unseen AUROC\n",
    "        y_pred = y_hat.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings_test.detach().clone().to('cpu').numpy()\n",
    "        perturbations['auc-unseen-surrogate'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "    # compute meta gradient\n",
    "    if 'meta' in algorithm:\n",
    "        meta_grad = torch.autograd.grad(loss, ratings)[0]\n",
    "\n",
    "    # define evaluation model\n",
    "    torch.manual_seed(seed_target)\n",
    "    eval_model = getattr(models, target)(n_users, n_items, n_factors_target, dropout_target).to(device)\n",
    "\n",
    "    # define optimizer and loss function for evaluation\n",
    "    if 'adam' in opt_target:\n",
    "        optimizer_eval = torch.optim.Adam(eval_model.parameters(), lr = lr_target)\n",
    "    else:\n",
    "        optimizer_eval = torch.optim.SGD(eval_model.parameters(), lr = lr_target)\n",
    "    loss_fn_eval = nn.BCELoss(reduction = 'mean')\n",
    "\n",
    "    # detach ratings and perturbs for eval model\n",
    "    ratings_eval = ratings.detach().clone()\n",
    "    perturbs_eval = perturbs.detach().clone()\n",
    "\n",
    "    # inner train  evaluation model\n",
    "    eval_model.train()\n",
    "    for i in range(T_target):\n",
    "        y_hat = eval_model(users, items).reshape(ratings_eval.shape)\n",
    "        loss_eval = loss_fn_eval(y_hat, ratings_eval)\n",
    "        optimizer_eval.zero_grad()\n",
    "        loss_eval.backward(retain_graph=retain_graph, create_graph=create_graph)\n",
    "        optimizer_eval.step()\n",
    "    eval_model.eval()\n",
    "\n",
    "    # compute and store accuracy of eval model after T training steps\n",
    "    with torch.no_grad():\n",
    "        # training accuracy and loss including perturbed edges\n",
    "        y_hat = eval_model(users, items).reshape(ratings_eval.shape)\n",
    "        perturbations['accuracy-before-target'].append(get_accuracy(y_hat, ratings_eval))\n",
    "        perturbations['loss-before-target'].append(loss_fn_eval(y_hat, ratings_eval).item())\n",
    "\n",
    "        # compute training AUROC including perturbed edges\n",
    "        y_pred = y_hat.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings_eval.detach().clone().to('cpu').numpy() \n",
    "        perturbations['auc-before-target'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "        # training accuracy and loss excluding perturbed edges\n",
    "        y_hat_masked = torch.masked_select(y_hat, perturbs_eval)\n",
    "        ratings_masked = torch.masked_select(ratings_eval, perturbs_eval)\n",
    "        perturbations['accuracy-after-target'].append(get_accuracy(y_hat_masked, ratings_masked))\n",
    "        perturbations['loss-after-target'].append(loss_fn_eval(y_hat_masked, ratings_masked).item())\n",
    "\n",
    "        # compute training AUROC excluding perturbed edges\n",
    "        y_pred = y_hat_masked.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings_masked.detach().clone().to('cpu').numpy()\n",
    "        perturbations['auc-after-target'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "    # compute and store accuracy of target model on unseen data\n",
    "    with torch.no_grad():\n",
    "        y_hat = eval_model(users_test, items_test).reshape(ratings_test.shape)\n",
    "        perturbations['accuracy-unseen-target'].append(get_accuracy(y_hat, ratings_test))\n",
    "        perturbations['loss-unseen-target'].append(loss_fn(y_hat, ratings_test).item())\n",
    "\n",
    "        # compute unseen AUROC\n",
    "        y_pred = y_hat.detach().clone().to('cpu').numpy()\n",
    "        y_actual = ratings_test.detach().clone().to('cpu').numpy()\n",
    "        perturbations['auc-unseen-target'].append(roc_auc_score(y_actual, y_pred))\n",
    "\n",
    "    # select best edge and perform perturbation\n",
    "    with torch.no_grad():\n",
    "        if 'meta' in algorithm:\n",
    "            mask = ratings.detach().int()\n",
    "            meta_grad[mask == 1] = 0\n",
    "            best_edge = meta_grad.argmax().item()\n",
    "            ratings[best_edge] = 1\n",
    "            perturbs[best_edge] = False\n",
    "\n",
    "            perturbations['edges'].append(best_edge)\n",
    "            perturbations['metagrad'].append(meta_grad[best_edge].item())\n",
    "\n",
    "        else:\n",
    "            best_edge = edges_to_perturb[delta]\n",
    "            ratings[best_edge] = 1 \n",
    "            perturbs[best_edge] = False\n",
    "\n",
    "            perturbations['edges'].append(best_edge)\n",
    "            perturbations['metagrad'].append(-1)\n",
    "\n",
    "sleep(1)\n",
    "# compute execution time\n",
    "exec_time = int(time.time() - start_time)\n",
    "exec_time = time.strftime(\"%Hh %Mm %Ss\", time.gmtime(exec_time))\n",
    "print('-> Execution time: {}'.format(exec_time))\n",
    "\n",
    "# process results\n",
    "perturbations = pd.DataFrame(perturbations)\n",
    "filename = 'extra-steam-{}({},{})-{}({},{})-{}-D={}-T({},{})-lr({},{})-seed({},{})-drop({},{})'.format(surrogate, n_factors_surrogate, opt_surrogate, target, n_factors_target, opt_target, algorithm, Delta, T_surrogate, T_target, lr_surrogate, lr_target, seed_surrogate, seed_target, dropout_surrogate, dropout_target)\n",
    "print('-> File name: {}'.format(filename))\n",
    "if save_results:\n",
    "    perturbations.to_csv('results/' + dataset + '/' + filename + '.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File name:  extra-steam-CFD(32,adam)-CFD(32,adam)-base-D=40000-T(50,50)-lr(0.3,0.3)-seed(0,1)-drop(0.2,0.2)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edges</th>\n",
       "      <th>metagrad</th>\n",
       "      <th>accuracy-before-surrogate</th>\n",
       "      <th>accuracy-after-surrogate</th>\n",
       "      <th>accuracy-unseen-surrogate</th>\n",
       "      <th>loss-before-surrogate</th>\n",
       "      <th>loss-after-surrogate</th>\n",
       "      <th>loss-unseen-surrogate</th>\n",
       "      <th>accuracy-before-target</th>\n",
       "      <th>accuracy-after-target</th>\n",
       "      <th>accuracy-unseen-target</th>\n",
       "      <th>loss-before-target</th>\n",
       "      <th>loss-after-target</th>\n",
       "      <th>loss-unseen-target</th>\n",
       "      <th>auc-before-surrogate</th>\n",
       "      <th>auc-after-surrogate</th>\n",
       "      <th>auc-unseen-surrogate</th>\n",
       "      <th>auc-before-target</th>\n",
       "      <th>auc-after-target</th>\n",
       "      <th>auc-unseen-target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>136339</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.928119</td>\n",
       "      <td>0.928119</td>\n",
       "      <td>0.928119</td>\n",
       "      <td>0.181197</td>\n",
       "      <td>0.181197</td>\n",
       "      <td>0.181197</td>\n",
       "      <td>0.927977</td>\n",
       "      <td>0.927977</td>\n",
       "      <td>0.927977</td>\n",
       "      <td>0.179583</td>\n",
       "      <td>0.179583</td>\n",
       "      <td>0.179583</td>\n",
       "      <td>0.982565</td>\n",
       "      <td>0.982565</td>\n",
       "      <td>0.982565</td>\n",
       "      <td>0.982907</td>\n",
       "      <td>0.982907</td>\n",
       "      <td>0.982907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>86848</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.928107</td>\n",
       "      <td>0.928107</td>\n",
       "      <td>0.928101</td>\n",
       "      <td>0.181196</td>\n",
       "      <td>0.181195</td>\n",
       "      <td>0.181203</td>\n",
       "      <td>0.927977</td>\n",
       "      <td>0.927982</td>\n",
       "      <td>0.927983</td>\n",
       "      <td>0.179629</td>\n",
       "      <td>0.179618</td>\n",
       "      <td>0.179618</td>\n",
       "      <td>0.982564</td>\n",
       "      <td>0.982564</td>\n",
       "      <td>0.982562</td>\n",
       "      <td>0.982894</td>\n",
       "      <td>0.982897</td>\n",
       "      <td>0.982897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>84965</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.927775</td>\n",
       "      <td>0.927774</td>\n",
       "      <td>0.927763</td>\n",
       "      <td>0.181196</td>\n",
       "      <td>0.181197</td>\n",
       "      <td>0.181263</td>\n",
       "      <td>0.928048</td>\n",
       "      <td>0.928053</td>\n",
       "      <td>0.928048</td>\n",
       "      <td>0.179638</td>\n",
       "      <td>0.179629</td>\n",
       "      <td>0.179662</td>\n",
       "      <td>0.982560</td>\n",
       "      <td>0.982560</td>\n",
       "      <td>0.982547</td>\n",
       "      <td>0.982891</td>\n",
       "      <td>0.982893</td>\n",
       "      <td>0.982885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>91552</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.927900</td>\n",
       "      <td>0.927904</td>\n",
       "      <td>0.927894</td>\n",
       "      <td>0.181222</td>\n",
       "      <td>0.181218</td>\n",
       "      <td>0.181287</td>\n",
       "      <td>0.928054</td>\n",
       "      <td>0.928064</td>\n",
       "      <td>0.928060</td>\n",
       "      <td>0.179668</td>\n",
       "      <td>0.179644</td>\n",
       "      <td>0.179676</td>\n",
       "      <td>0.982556</td>\n",
       "      <td>0.982557</td>\n",
       "      <td>0.982544</td>\n",
       "      <td>0.982884</td>\n",
       "      <td>0.982891</td>\n",
       "      <td>0.982883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>42170</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.927923</td>\n",
       "      <td>0.927928</td>\n",
       "      <td>0.927912</td>\n",
       "      <td>0.181202</td>\n",
       "      <td>0.181198</td>\n",
       "      <td>0.181272</td>\n",
       "      <td>0.927983</td>\n",
       "      <td>0.927993</td>\n",
       "      <td>0.927983</td>\n",
       "      <td>0.179660</td>\n",
       "      <td>0.179635</td>\n",
       "      <td>0.179676</td>\n",
       "      <td>0.982561</td>\n",
       "      <td>0.982562</td>\n",
       "      <td>0.982547</td>\n",
       "      <td>0.982884</td>\n",
       "      <td>0.982891</td>\n",
       "      <td>0.982880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39995</th>\n",
       "      <td>120889</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.904784</td>\n",
       "      <td>0.904688</td>\n",
       "      <td>0.751735</td>\n",
       "      <td>0.227740</td>\n",
       "      <td>0.224929</td>\n",
       "      <td>0.980158</td>\n",
       "      <td>0.905465</td>\n",
       "      <td>0.905702</td>\n",
       "      <td>0.752701</td>\n",
       "      <td>0.226556</td>\n",
       "      <td>0.222966</td>\n",
       "      <td>0.977310</td>\n",
       "      <td>0.965780</td>\n",
       "      <td>0.975320</td>\n",
       "      <td>0.842409</td>\n",
       "      <td>0.966319</td>\n",
       "      <td>0.976022</td>\n",
       "      <td>0.843149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39996</th>\n",
       "      <td>1290</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.904784</td>\n",
       "      <td>0.904688</td>\n",
       "      <td>0.751735</td>\n",
       "      <td>0.227740</td>\n",
       "      <td>0.224929</td>\n",
       "      <td>0.980158</td>\n",
       "      <td>0.905465</td>\n",
       "      <td>0.905702</td>\n",
       "      <td>0.752701</td>\n",
       "      <td>0.226556</td>\n",
       "      <td>0.222966</td>\n",
       "      <td>0.977310</td>\n",
       "      <td>0.965780</td>\n",
       "      <td>0.975320</td>\n",
       "      <td>0.842409</td>\n",
       "      <td>0.966319</td>\n",
       "      <td>0.976022</td>\n",
       "      <td>0.843149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39997</th>\n",
       "      <td>93150</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.904784</td>\n",
       "      <td>0.904688</td>\n",
       "      <td>0.751735</td>\n",
       "      <td>0.227740</td>\n",
       "      <td>0.224929</td>\n",
       "      <td>0.980158</td>\n",
       "      <td>0.905465</td>\n",
       "      <td>0.905702</td>\n",
       "      <td>0.752701</td>\n",
       "      <td>0.226556</td>\n",
       "      <td>0.222966</td>\n",
       "      <td>0.977310</td>\n",
       "      <td>0.965780</td>\n",
       "      <td>0.975320</td>\n",
       "      <td>0.842409</td>\n",
       "      <td>0.966319</td>\n",
       "      <td>0.976022</td>\n",
       "      <td>0.843149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39998</th>\n",
       "      <td>12125</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.904784</td>\n",
       "      <td>0.904688</td>\n",
       "      <td>0.751735</td>\n",
       "      <td>0.227740</td>\n",
       "      <td>0.224929</td>\n",
       "      <td>0.980158</td>\n",
       "      <td>0.905465</td>\n",
       "      <td>0.905702</td>\n",
       "      <td>0.752701</td>\n",
       "      <td>0.226556</td>\n",
       "      <td>0.222966</td>\n",
       "      <td>0.977310</td>\n",
       "      <td>0.965780</td>\n",
       "      <td>0.975320</td>\n",
       "      <td>0.842409</td>\n",
       "      <td>0.966319</td>\n",
       "      <td>0.976022</td>\n",
       "      <td>0.843149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39999</th>\n",
       "      <td>152852</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.904784</td>\n",
       "      <td>0.904688</td>\n",
       "      <td>0.751735</td>\n",
       "      <td>0.227740</td>\n",
       "      <td>0.224929</td>\n",
       "      <td>0.980158</td>\n",
       "      <td>0.905465</td>\n",
       "      <td>0.905702</td>\n",
       "      <td>0.752701</td>\n",
       "      <td>0.226556</td>\n",
       "      <td>0.222966</td>\n",
       "      <td>0.977310</td>\n",
       "      <td>0.965780</td>\n",
       "      <td>0.975320</td>\n",
       "      <td>0.842409</td>\n",
       "      <td>0.966319</td>\n",
       "      <td>0.976022</td>\n",
       "      <td>0.843149</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>40000 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        edges  metagrad  accuracy-before-surrogate  accuracy-after-surrogate  \\\n",
       "0      136339        -1                   0.928119                  0.928119   \n",
       "1       86848        -1                   0.928107                  0.928107   \n",
       "2       84965        -1                   0.927775                  0.927774   \n",
       "3       91552        -1                   0.927900                  0.927904   \n",
       "4       42170        -1                   0.927923                  0.927928   \n",
       "...       ...       ...                        ...                       ...   \n",
       "39995  120889        -1                   0.904784                  0.904688   \n",
       "39996    1290        -1                   0.904784                  0.904688   \n",
       "39997   93150        -1                   0.904784                  0.904688   \n",
       "39998   12125        -1                   0.904784                  0.904688   \n",
       "39999  152852        -1                   0.904784                  0.904688   \n",
       "\n",
       "       accuracy-unseen-surrogate  loss-before-surrogate  loss-after-surrogate  \\\n",
       "0                       0.928119               0.181197              0.181197   \n",
       "1                       0.928101               0.181196              0.181195   \n",
       "2                       0.927763               0.181196              0.181197   \n",
       "3                       0.927894               0.181222              0.181218   \n",
       "4                       0.927912               0.181202              0.181198   \n",
       "...                          ...                    ...                   ...   \n",
       "39995                   0.751735               0.227740              0.224929   \n",
       "39996                   0.751735               0.227740              0.224929   \n",
       "39997                   0.751735               0.227740              0.224929   \n",
       "39998                   0.751735               0.227740              0.224929   \n",
       "39999                   0.751735               0.227740              0.224929   \n",
       "\n",
       "       loss-unseen-surrogate  accuracy-before-target  accuracy-after-target  \\\n",
       "0                   0.181197                0.927977               0.927977   \n",
       "1                   0.181203                0.927977               0.927982   \n",
       "2                   0.181263                0.928048               0.928053   \n",
       "3                   0.181287                0.928054               0.928064   \n",
       "4                   0.181272                0.927983               0.927993   \n",
       "...                      ...                     ...                    ...   \n",
       "39995               0.980158                0.905465               0.905702   \n",
       "39996               0.980158                0.905465               0.905702   \n",
       "39997               0.980158                0.905465               0.905702   \n",
       "39998               0.980158                0.905465               0.905702   \n",
       "39999               0.980158                0.905465               0.905702   \n",
       "\n",
       "       accuracy-unseen-target  loss-before-target  loss-after-target  \\\n",
       "0                    0.927977            0.179583           0.179583   \n",
       "1                    0.927983            0.179629           0.179618   \n",
       "2                    0.928048            0.179638           0.179629   \n",
       "3                    0.928060            0.179668           0.179644   \n",
       "4                    0.927983            0.179660           0.179635   \n",
       "...                       ...                 ...                ...   \n",
       "39995                0.752701            0.226556           0.222966   \n",
       "39996                0.752701            0.226556           0.222966   \n",
       "39997                0.752701            0.226556           0.222966   \n",
       "39998                0.752701            0.226556           0.222966   \n",
       "39999                0.752701            0.226556           0.222966   \n",
       "\n",
       "       loss-unseen-target  auc-before-surrogate  auc-after-surrogate  \\\n",
       "0                0.179583              0.982565             0.982565   \n",
       "1                0.179618              0.982564             0.982564   \n",
       "2                0.179662              0.982560             0.982560   \n",
       "3                0.179676              0.982556             0.982557   \n",
       "4                0.179676              0.982561             0.982562   \n",
       "...                   ...                   ...                  ...   \n",
       "39995            0.977310              0.965780             0.975320   \n",
       "39996            0.977310              0.965780             0.975320   \n",
       "39997            0.977310              0.965780             0.975320   \n",
       "39998            0.977310              0.965780             0.975320   \n",
       "39999            0.977310              0.965780             0.975320   \n",
       "\n",
       "       auc-unseen-surrogate  auc-before-target  auc-after-target  \\\n",
       "0                  0.982565           0.982907          0.982907   \n",
       "1                  0.982562           0.982894          0.982897   \n",
       "2                  0.982547           0.982891          0.982893   \n",
       "3                  0.982544           0.982884          0.982891   \n",
       "4                  0.982547           0.982884          0.982891   \n",
       "...                     ...                ...               ...   \n",
       "39995              0.842409           0.966319          0.976022   \n",
       "39996              0.842409           0.966319          0.976022   \n",
       "39997              0.842409           0.966319          0.976022   \n",
       "39998              0.842409           0.966319          0.976022   \n",
       "39999              0.842409           0.966319          0.976022   \n",
       "\n",
       "       auc-unseen-target  \n",
       "0               0.982907  \n",
       "1               0.982897  \n",
       "2               0.982885  \n",
       "3               0.982883  \n",
       "4               0.982880  \n",
       "...                  ...  \n",
       "39995           0.843149  \n",
       "39996           0.843149  \n",
       "39997           0.843149  \n",
       "39998           0.843149  \n",
       "39999           0.843149  \n",
       "\n",
       "[40000 rows x 20 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('File name: ', filename)\n",
    "if 'meta' in algorithm:\n",
    "    print('Max meta-gradient: ', perturbations['metagrad'].max())\n",
    "perturbations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Plots**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File name:  extra-steam-CFD(32,adam)-CFD(32,adam)-base-D=40000-T(50,50)-lr(0.3,0.3)-seed(0,1)-drop(0.2,0.2)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Process perturbations data\n",
    "perturbations_reset = perturbations.reset_index()\n",
    "perturbations_reset.columns = ['# Perturbations', 'Perturbed edge', 'Meta gradient', 'Accuracy including perturbed edges (surrogate)', 'Accuracy excluding perturbed edges (surrogate)', 'Accuracy unseen (surrogate)', 'Loss including perturbed edges (surrogate)', 'Loss excluding perturbed edges (surrogate)', 'Loss unseen (surrogate)', 'Accuracy including perturbed edges (target)', 'Accuracy excluding perturbed edges (target)', 'Accuracy unseen (target)', 'Loss including perturbed edges (target)', 'Loss excluding perturbed edges (target)', 'Loss unseen (target)', 'AUROC including perturbed edges (surrogate)', 'AUROC excluding perturbed edges (surrogate)', 'AUROC unseen (surrogate)', 'AUROC including perturbed edges (target)', 'AUROC excluding perturbed edges (target)', 'AUROC unseen (target)']\n",
    "perturbations_reset\n",
    "\n",
    "# Figure specifications\n",
    "sns.set_theme()\n",
    "figsize = (7, 6)\n",
    "print('File name: ', filename)\n",
    "\n",
    "# Loss on training data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Loss excluding perturbed edges (surrogate)', color = 'blue').set_title('Comparison of loss on training data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Loss excluding perturbed edges (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(labels = ['Loss (surrogate)', 'Loss (target)'])\n",
    "if save_results: \n",
    "    fig.savefig('plots/' + dataset + '/losses-' + filename + '.png')\n",
    "plt.show()\n",
    "\n",
    "# Loss on unseen data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Loss unseen (surrogate)', color = 'blue').set_title('Comparison of loss on unseen data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Loss unseen (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(labels = ['Loss unseen (surrogate)', 'Loss unseen (target)'])\n",
    "if save_results: \n",
    "    fig.savefig('plots/' + dataset + '/losses-unseen-' + filename + '.png')\n",
    "plt.show()\n",
    "\n",
    "# Accuracy on training data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Accuracy excluding perturbed edges (surrogate)', color = 'blue').set_title('Comparison of accuracy on training data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Accuracy excluding perturbed edges (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(labels = ['Accuracy (surrogate)', 'Accuracy (target)'])\n",
    "if save_results:\n",
    "    fig.savefig('plots/' + dataset + '/accuracy-' + filename + '.png')\n",
    "plt.show()\n",
    "\n",
    "# Accuracy on unseen data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Accuracy unseen (surrogate)', color = 'blue').set_title('Comparison of accuracy on unseen data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'Accuracy unseen (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(labels = ['Accuracy unseen (surrogate)', 'Accuracy unseen (target)'])\n",
    "if save_results: \n",
    "    fig.savefig('plots/' + dataset + '/accuracy-unseen-' + filename + '.png')\n",
    "plt.show()\n",
    "\n",
    "# AUROC on training data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'AUROC excluding perturbed edges (surrogate)', color = 'blue').set_title('Comparison of AUROC on training data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'AUROC excluding perturbed edges (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('AUROC')\n",
    "plt.legend(labels = ['AUROC (surrogate)', 'AUROC (target)'])\n",
    "if save_results:\n",
    "    fig.savefig('plots/' + dataset + '/AUROC-' + filename + '.png')\n",
    "plt.show()\n",
    "\n",
    "# AUROC on unseen data plot\n",
    "plt.figure(figsize = figsize)\n",
    "sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'AUROC unseen (surrogate)', color = 'blue').set_title('Comparison of AUROC on unseen data ({})'.format(algorithm))\n",
    "fig = sns.lineplot(data = perturbations_reset, x = '# Perturbations', y = 'AUROC unseen (target)', color = 'green').get_figure()\n",
    "plt.xlabel('Number of perturbations')\n",
    "plt.ylabel('AUROC')\n",
    "plt.legend(labels = ['AUROC unseen (surrogate)', 'AUROC unseen (target)'])\n",
    "if save_results:\n",
    "    fig.savefig('plots/' + dataset + '/AUROC-unseen-' + filename + '.png')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "6f38835821387ecea7238337192aa99e87ed1a9c9c1fa6562e207de7e0c31193"
  },
  "kernelspec": {
   "display_name": "Python 3.7.11 ('PyG': conda)",
   "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"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}