toy-sine/main-sine-CIDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# install / upgrade necessary packages\n",
    "# !pip install easydict\n",
    "# !pip install numpy\n",
    "# !pip install numpy --upgrade # upgrade numpy to at least 1.19\n",
    "# !pip install torch --upgrade # upgrade pytorch to 1.5.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from easydict import EasyDict\n",
    "args = EasyDict()\n",
    "\n",
    "args.epochs=100\n",
    "args.dropout=0.0\n",
    "args.lr=5e-5\n",
    "args.gamma_exp=1000\n",
    "args.hidden=800\n",
    "args.ratio=1\n",
    "args.dis_lambda=1.0\n",
    "args.lambda_m=0.0\n",
    "args.wgan='wgan'\n",
    "args.clamp_lower=-0.15\n",
    "args.clamp_upper=0.15\n",
    "args.batch_size=100\n",
    "args.num_train=100\n",
    "args.loss='default'\n",
    "args.evaluate=False\n",
    "args.checkpoint='none'\n",
    "args.save_head='tmp'\n",
    "args.save_interval=20\n",
    "args.log_interval=20\n",
    "args.log_file='tmp_mlp'\n",
    "args.seed=2 #2\n",
    "args.cuda=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Numpy version: 1.19.0\n",
      "Pytorch version: 1.5.1+cu101\n"
     ]
    }
   ],
   "source": [
    "from __future__ import print_function\n",
    "from torch.utils.data.sampler import SubsetRandomSampler\n",
    "import numpy as np\n",
    "import random\n",
    "import argparse\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "import torch.optim.lr_scheduler as lr_scheduler\n",
    "from torchvision import datasets, transforms\n",
    "from torch.autograd import Variable\n",
    "from utils import plain_log\n",
    "from utils import write_pickle,read_pickle\n",
    "from utils import masked_cross_entropy\n",
    "from utils import gaussian_loss\n",
    "from torch.utils import data\n",
    "from data_loader import toydata\n",
    "\n",
    "label_noise_std = 0.20\n",
    "use_label_noise = False\n",
    "use_inverse_weighted = True\n",
    "discr_thres = 999.999\n",
    "normalize = True\n",
    "train_discr_step_tot = 2\n",
    "train_discr_step_extra = 0\n",
    "slow_lrD_decay = 1\n",
    "norm = 8\n",
    "fname_save = 'pred_tmp.pkl'\n",
    "fname = 'sine.pkl'\n",
    "train_list = list(range(12))\n",
    "mask_list = [1] * 5 + [0] * 7\n",
    "test_list = list(range(12))\n",
    "\n",
    "torch.manual_seed(args.seed)\n",
    "if args.cuda:\n",
    "    torch.cuda.manual_seed(args.seed)\n",
    "random.seed(args.seed)\n",
    "torch.manual_seed(int(args.seed))\n",
    "\n",
    "kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {}\n",
    "train_loader = torch.utils.data.DataLoader(\n",
    "    toydata(fname, train_list, normalize, mask_list),\n",
    "    shuffle=True,\n",
    "    batch_size=args.batch_size, **kwargs)\n",
    "test_loader = torch.utils.data.DataLoader(\n",
    "    toydata(fname, test_list, normalize),\n",
    "    batch_size=args.batch_size, **kwargs)\n",
    "\n",
    "print(\"Numpy version:\", np.__version__)\n",
    "print(\"Pytorch version:\", torch.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x324 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# load raw data and plot it for visualization only\n",
    "from plot import plot_dataset\n",
    "import matplotlib.pyplot as plt\n",
    "data_pkl = read_pickle(f'./sine.pkl')\n",
    "plot_dataset(data_pkl)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Encoder: domain as input to generate feature, and then concatenate, deep dynamic layers\n",
    "class DomainEnc(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(DomainEnc, self).__init__()\n",
    "        self.hidden = args.hidden\n",
    "        self.ratio = float(args.ratio)\n",
    "        self.dropout = args.dropout\n",
    "\n",
    "        self.fc1 = nn.Linear(2, self.hidden)\n",
    "        self.drop1 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc2 = nn.Linear(self.hidden + int(self.hidden // self.ratio), self.hidden + int(self.hidden // self.ratio))\n",
    "        self.drop2 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc3 = nn.Linear(self.hidden + int(self.hidden // self.ratio), self.hidden + int(self.hidden // self.ratio))\n",
    "        self.drop3 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc4 = nn.Linear(self.hidden + int(self.hidden // self.ratio), self.hidden + int(self.hidden // self.ratio))\n",
    "        self.drop4 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc_final = nn.Linear(self.hidden + int(self.hidden // self.ratio), self.hidden)\n",
    "\n",
    "        self.fc1_var = nn.Linear(1, int(self.hidden // self.ratio))\n",
    "        self.fc2_var = nn.Linear(int(self.hidden // self.ratio), int(self.hidden // self.ratio))\n",
    "        self.fc3_var = nn.Linear(int(self.hidden // self.ratio), int(self.hidden // self.ratio))\n",
    "        self.drop1_var = nn.Dropout(self.dropout)\n",
    "        self.drop2_var = nn.Dropout(self.dropout)\n",
    "        self.drop3_var = nn.Dropout(self.dropout)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x, domain = x\n",
    "        domain = domain.unsqueeze(1) / norm\n",
    "\n",
    "        # side branch for variable FC\n",
    "        x_domain = F.relu(self.fc1_var(domain))\n",
    "        x_domain = self.drop1_var(x_domain)\n",
    "\n",
    "        # main branch\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = self.drop1(x)\n",
    "\n",
    "        # combine feature in the middle\n",
    "        x = torch.cat((x, x_domain), dim=1)\n",
    "\n",
    "        x = F.relu(self.fc2(x))\n",
    "        x = self.drop2(x)\n",
    "\n",
    "        x = F.relu(self.fc3(x))\n",
    "        x = self.drop3(x)\n",
    "\n",
    "        x = F.relu(self.fc4(x))\n",
    "        x = self.drop4(x)\n",
    "\n",
    "        # continue main branch\n",
    "        x = F.relu(self.fc_final(x))\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predictor\n",
    "class DomainPred(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(DomainPred, self).__init__()\n",
    "        self.hidden = args.hidden\n",
    "        self.dropout = args.dropout\n",
    "\n",
    "        self.drop0 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc1 = nn.Linear(self.hidden, self.hidden)\n",
    "        self.drop1 = nn.Dropout(self.dropout)\n",
    "\n",
    "\n",
    "        self.fc_final = nn.Linear(self.hidden, 2)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x, domain = x\n",
    "        domain = domain.unsqueeze(1) / norm\n",
    "\n",
    "        x = self.drop0(x)\n",
    "\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = self.drop1(x)\n",
    "\n",
    "        x = self.fc_final(x)\n",
    "        \n",
    "        return F.log_softmax(x, dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Discriminator: with BN layers after each FC, dual output\n",
    "class DomainDDisc(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(DomainDDisc, self).__init__()\n",
    "        self.hidden = args.hidden\n",
    "        self.dropout = args.dropout\n",
    "\n",
    "        self.drop2 = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc3_m = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn3_m = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop3_m = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc3_s = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn3_s = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop3_s = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc4_m = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn4_m = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop4_m = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc4_s = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn4_s = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop4_s = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc5_m = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn5_m = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop5_m = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc5_s = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn5_s = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop5_s = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc6_m = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn6_m = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop6_m = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc6_s = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn6_s = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop6_s = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc7_m = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn7_m = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop7_m = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc7_s = nn.Linear(self.hidden, self.hidden)\n",
    "        self.bn7_s = nn.BatchNorm1d(self.hidden)\n",
    "        self.drop7_s = nn.Dropout(self.dropout)\n",
    "\n",
    "        self.fc_final_m = nn.Linear(self.hidden, 1)\n",
    "        self.fc_final_s = nn.Linear(self.hidden, 1)        \n",
    "\n",
    "    def forward(self, x):\n",
    "        x, domain = x\n",
    "        domain = domain.unsqueeze(1) / norm\n",
    "\n",
    "        x = self.drop2(x)\n",
    "\n",
    "        x_m = F.relu(self.bn3_m(self.fc3_m(x)))\n",
    "        x_m = self.drop3_m(x_m)\n",
    "\n",
    "        x_s = F.relu(self.bn3_s(self.fc3_s(x)))\n",
    "        x_s = self.drop3_s(x_s)\n",
    "\n",
    "        x_m = F.relu(self.bn4_m(self.fc4_m(x_m)))\n",
    "        x_m = self.drop4_m(x_m)\n",
    "\n",
    "        x_s = F.relu(self.bn4_s(self.fc4_s(x_s)))\n",
    "        x_s = self.drop4_s(x_s)\n",
    "\n",
    "        x_m = F.relu(self.bn5_m(self.fc5_m(x_m)))\n",
    "        x_m = self.drop5_m(x_m)\n",
    "\n",
    "        x_s = F.relu(self.bn5_s(self.fc5_s(x_s)))\n",
    "        x_s = self.drop5_s(x_s)\n",
    "\n",
    "        x_m = F.relu(self.bn6_m(self.fc6_m(x_m)))\n",
    "        x_m = self.drop6_m(x_m)\n",
    "\n",
    "        x_s = F.relu(self.bn6_s(self.fc6_s(x_s)))\n",
    "        x_s = self.drop6_s(x_s)\n",
    "\n",
    "        x_m = F.relu(self.bn7_m(self.fc7_m(x_m)))\n",
    "        x_m = self.drop7_m(x_m)\n",
    "\n",
    "        x_s = F.relu(self.bn7_s(self.fc7_s(x_s)))\n",
    "        x_s = self.drop7_s(x_s)\n",
    "\n",
    "        x_m = self.fc_final_m(x_m)\n",
    "        x_s = self.fc_final_s(x_s) # log sigma^2\n",
    "\n",
    "        return (x_m, x_s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create models\n",
    "encoder = DomainEnc()\n",
    "predictor = DomainPred()\n",
    "discriminator = DomainDDisc()\n",
    "models = [encoder, predictor, discriminator]\n",
    "if args.cuda:\n",
    "    for model in models:\n",
    "        model.cuda()    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Epoch 1: avg_discr_loss = 0.38046, avg_pred_loss = 0.697, avg_total_loss = 0.297\n",
      "Test set: Average loss: 0.6868983, Accuracy: 1776/2280 (77.89%)\n",
      "Train Epoch 2: avg_discr_loss = 0.17093, avg_pred_loss = 0.685, avg_total_loss = 0.471\n",
      "Test set: Average loss: 0.6679511, Accuracy: 1994/2280 (87.46%)\n",
      "Train Epoch 3: avg_discr_loss = 0.02658, avg_pred_loss = 0.648, avg_total_loss = 0.560\n",
      "Test set: Average loss: 0.6036035, Accuracy: 2053/2280 (90.04%)\n",
      "Train Epoch 4: avg_discr_loss = -0.10381, avg_pred_loss = 0.518, avg_total_loss = 0.547\n",
      "Test set: Average loss: 0.3850066, Accuracy: 2079/2280 (91.18%)\n",
      "Train Epoch 5: avg_discr_loss = -0.22480, avg_pred_loss = 0.290, avg_total_loss = 0.431\n",
      "Test set: Average loss: 0.3233958, Accuracy: 1956/2280 (85.79%)\n",
      "Train Epoch 6: avg_discr_loss = -0.31789, avg_pred_loss = 0.241, avg_total_loss = 0.470\n",
      "Test set: Average loss: 0.2282375, Accuracy: 2064/2280 (90.53%)\n",
      "Train Epoch 7: avg_discr_loss = -0.40245, avg_pred_loss = 0.233, avg_total_loss = 0.543\n",
      "Test set: Average loss: 0.2218932, Accuracy: 2074/2280 (90.96%)\n",
      "Train Epoch 8: avg_discr_loss = -0.47270, avg_pred_loss = 0.247, avg_total_loss = 0.633\n",
      "Test set: Average loss: 0.2753940, Accuracy: 2010/2280 (88.16%)\n",
      "Train Epoch 9: avg_discr_loss = -0.51825, avg_pred_loss = 0.256, avg_total_loss = 0.700\n",
      "Test set: Average loss: 0.2337762, Accuracy: 2054/2280 (90.09%)\n",
      "Train Epoch 10: avg_discr_loss = -0.59865, avg_pred_loss = 0.228, avg_total_loss = 0.753\n",
      "Test set: Average loss: 0.2284740, Accuracy: 2056/2280 (90.18%)\n",
      "Train Epoch 11: avg_discr_loss = -0.66126, avg_pred_loss = 0.228, avg_total_loss = 0.818\n",
      "Test set: Average loss: 0.2339884, Accuracy: 2056/2280 (90.18%)\n",
      "Train Epoch 12: avg_discr_loss = -0.69595, avg_pred_loss = 0.231, avg_total_loss = 0.858\n",
      "Test set: Average loss: 0.2277895, Accuracy: 2059/2280 (90.31%)\n",
      "Train Epoch 13: avg_discr_loss = -0.69572, avg_pred_loss = 0.259, avg_total_loss = 0.891\n",
      "Test set: Average loss: 0.2435271, Accuracy: 2037/2280 (89.34%)\n",
      "Train Epoch 14: avg_discr_loss = -0.70526, avg_pred_loss = 0.244, avg_total_loss = 0.897\n",
      "Test set: Average loss: 0.2336632, Accuracy: 2051/2280 (89.96%)\n",
      "Train Epoch 15: avg_discr_loss = -0.72856, avg_pred_loss = 0.237, avg_total_loss = 0.902\n",
      "Test set: Average loss: 0.2341053, Accuracy: 2049/2280 (89.87%)\n",
      "Train Epoch 16: avg_discr_loss = -0.71738, avg_pred_loss = 0.242, avg_total_loss = 0.901\n",
      "Test set: Average loss: 0.2342037, Accuracy: 2053/2280 (90.04%)\n",
      "Train Epoch 17: avg_discr_loss = -0.69086, avg_pred_loss = 0.223, avg_total_loss = 0.867\n",
      "Test set: Average loss: 0.2370743, Accuracy: 2046/2280 (89.74%)\n",
      "Train Epoch 18: avg_discr_loss = -0.50879, avg_pred_loss = 0.224, avg_total_loss = 0.747\n",
      "Test set: Average loss: 0.2257361, Accuracy: 2055/2280 (90.13%)\n",
      "Train Epoch 19: avg_discr_loss = -0.54489, avg_pred_loss = 0.224, avg_total_loss = 0.748\n",
      "Test set: Average loss: 0.2342282, Accuracy: 2046/2280 (89.74%)\n",
      "Train Epoch 20: avg_discr_loss = -0.35881, avg_pred_loss = 0.230, avg_total_loss = 0.594\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hehaodele/anaconda3/lib/python3.7/site-packages/torch/serialization.py:402: UserWarning: Couldn't retrieve source code for container of type DomainEnc. It won't be checked for correctness upon loading.\n",
      "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
      "/home/hehaodele/anaconda3/lib/python3.7/site-packages/torch/serialization.py:402: UserWarning: Couldn't retrieve source code for container of type DomainPred. It won't be checked for correctness upon loading.\n",
      "  \"type \" + obj.__name__ + \". It won't be checked \"\n",
      "/home/hehaodele/anaconda3/lib/python3.7/site-packages/torch/serialization.py:402: UserWarning: Couldn't retrieve source code for container of type DomainDDisc. It won't be checked for correctness upon loading.\n",
      "  \"type \" + obj.__name__ + \". It won't be checked \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set: Average loss: 0.2341022, Accuracy: 2049/2280 (89.87%)\n",
      "Train Epoch 21: avg_discr_loss = -0.20351, avg_pred_loss = 0.219, avg_total_loss = 0.337\n",
      "Test set: Average loss: 0.2277152, Accuracy: 2058/2280 (90.26%)\n",
      "Train Epoch 22: avg_discr_loss = -0.43588, avg_pred_loss = 0.221, avg_total_loss = 0.606\n",
      "Test set: Average loss: 0.2334848, Accuracy: 2055/2280 (90.13%)\n",
      "Train Epoch 23: avg_discr_loss = -0.53396, avg_pred_loss = 0.220, avg_total_loss = 0.722\n",
      "Test set: Average loss: 0.2345715, Accuracy: 2052/2280 (90.00%)\n",
      "Train Epoch 24: avg_discr_loss = -0.60007, avg_pred_loss = 0.214, avg_total_loss = 0.767\n",
      "Test set: Average loss: 0.2291436, Accuracy: 2047/2280 (89.78%)\n",
      "Train Epoch 25: avg_discr_loss = -0.59954, avg_pred_loss = 0.247, avg_total_loss = 0.813\n",
      "Test set: Average loss: 0.2611234, Accuracy: 2038/2280 (89.39%)\n",
      "Train Epoch 26: avg_discr_loss = -0.64385, avg_pred_loss = 0.233, avg_total_loss = 0.839\n",
      "Test set: Average loss: 0.2509042, Accuracy: 2038/2280 (89.39%)\n",
      "Train Epoch 27: avg_discr_loss = -0.56868, avg_pred_loss = 0.211, avg_total_loss = 0.749\n",
      "Test set: Average loss: 0.2426790, Accuracy: 2052/2280 (90.00%)\n",
      "Train Epoch 28: avg_discr_loss = -0.30839, avg_pred_loss = 0.208, avg_total_loss = 0.482\n",
      "Test set: Average loss: 0.2330124, Accuracy: 2055/2280 (90.13%)\n",
      "Train Epoch 29: avg_discr_loss = -0.59567, avg_pred_loss = 0.209, avg_total_loss = 0.750\n",
      "Test set: Average loss: 0.2315386, Accuracy: 2048/2280 (89.82%)\n",
      "Train Epoch 30: avg_discr_loss = -0.54131, avg_pred_loss = 0.201, avg_total_loss = 0.725\n",
      "Test set: Average loss: 0.2667079, Accuracy: 2039/2280 (89.43%)\n",
      "Train Epoch 31: avg_discr_loss = -0.49911, avg_pred_loss = 0.206, avg_total_loss = 0.684\n",
      "Test set: Average loss: 0.2322317, Accuracy: 2051/2280 (89.96%)\n",
      "Train Epoch 32: avg_discr_loss = -0.22903, avg_pred_loss = 0.214, avg_total_loss = 0.350\n",
      "Test set: Average loss: 0.3062382, Accuracy: 1991/2280 (87.32%)\n",
      "Train Epoch 33: avg_discr_loss = -0.46593, avg_pred_loss = 0.216, avg_total_loss = 0.633\n",
      "Test set: Average loss: 0.2506123, Accuracy: 2048/2280 (89.82%)\n",
      "Train Epoch 34: avg_discr_loss = -0.62480, avg_pred_loss = 0.197, avg_total_loss = 0.762\n",
      "Test set: Average loss: 0.3030595, Accuracy: 2011/2280 (88.20%)\n",
      "Train Epoch 35: avg_discr_loss = -0.60736, avg_pred_loss = 0.202, avg_total_loss = 0.777\n",
      "Test set: Average loss: 0.2291592, Accuracy: 2054/2280 (90.09%)\n",
      "Train Epoch 36: avg_discr_loss = -0.55989, avg_pred_loss = 0.214, avg_total_loss = 0.735\n",
      "Test set: Average loss: 0.2426482, Accuracy: 2049/2280 (89.87%)\n",
      "Train Epoch 37: avg_discr_loss = -0.17896, avg_pred_loss = 0.204, avg_total_loss = 0.341\n",
      "Test set: Average loss: 0.2255306, Accuracy: 2063/2280 (90.48%)\n",
      "Train Epoch 38: avg_discr_loss = -0.54503, avg_pred_loss = 0.193, avg_total_loss = 0.670\n",
      "Test set: Average loss: 0.2404009, Accuracy: 2050/2280 (89.91%)\n",
      "Train Epoch 39: avg_discr_loss = -0.58716, avg_pred_loss = 0.182, avg_total_loss = 0.716\n",
      "Test set: Average loss: 0.2404049, Accuracy: 2059/2280 (90.31%)\n",
      "Train Epoch 40: avg_discr_loss = -0.53420, avg_pred_loss = 0.189, avg_total_loss = 0.699\n",
      "Test set: Average loss: 0.2721294, Accuracy: 2047/2280 (89.78%)\n",
      "Train Epoch 41: avg_discr_loss = -0.13437, avg_pred_loss = 0.176, avg_total_loss = 0.247\n",
      "Test set: Average loss: 0.2440502, Accuracy: 2064/2280 (90.53%)\n",
      "Train Epoch 42: avg_discr_loss = -0.53033, avg_pred_loss = 0.181, avg_total_loss = 0.617\n",
      "Test set: Average loss: 0.2389245, Accuracy: 2059/2280 (90.31%)\n",
      "Train Epoch 43: avg_discr_loss = -0.65872, avg_pred_loss = 0.170, avg_total_loss = 0.769\n",
      "Test set: Average loss: 0.2329717, Accuracy: 2065/2280 (90.57%)\n",
      "Train Epoch 44: avg_discr_loss = -0.34581, avg_pred_loss = 0.160, avg_total_loss = 0.473\n",
      "Test set: Average loss: 0.2495045, Accuracy: 2082/2280 (91.32%)\n",
      "Train Epoch 45: avg_discr_loss = -0.50854, avg_pred_loss = 0.179, avg_total_loss = 0.620\n",
      "Test set: Average loss: 0.2270802, Accuracy: 2091/2280 (91.71%)\n",
      "Train Epoch 46: avg_discr_loss = -0.19858, avg_pred_loss = 0.149, avg_total_loss = 0.291\n",
      "Test set: Average loss: 0.2428598, Accuracy: 2074/2280 (90.96%)\n",
      "Train Epoch 47: avg_discr_loss = -0.40790, avg_pred_loss = 0.149, avg_total_loss = 0.473\n",
      "Test set: Average loss: 0.2266076, Accuracy: 2077/2280 (91.10%)\n",
      "Train Epoch 48: avg_discr_loss = -0.38420, avg_pred_loss = 0.150, avg_total_loss = 0.474\n",
      "Test set: Average loss: 0.2347171, Accuracy: 2076/2280 (91.05%)\n",
      "Train Epoch 49: avg_discr_loss = -0.60060, avg_pred_loss = 0.150, avg_total_loss = 0.694\n",
      "Test set: Average loss: 0.2803703, Accuracy: 2052/2280 (90.00%)\n",
      "Train Epoch 50: avg_discr_loss = -0.45592, avg_pred_loss = 0.164, avg_total_loss = 0.590\n",
      "Test set: Average loss: 0.2530317, Accuracy: 2073/2280 (90.92%)\n",
      "Train Epoch 51: avg_discr_loss = -0.11345, avg_pred_loss = 0.149, avg_total_loss = 0.174\n",
      "Test set: Average loss: 0.2438038, Accuracy: 2083/2280 (91.36%)\n",
      "Train Epoch 52: avg_discr_loss = -0.37225, avg_pred_loss = 0.144, avg_total_loss = 0.417\n",
      "Test set: Average loss: 0.2450541, Accuracy: 2068/2280 (90.70%)\n",
      "Train Epoch 53: avg_discr_loss = -0.36234, avg_pred_loss = 0.144, avg_total_loss = 0.454\n",
      "Test set: Average loss: 0.2563790, Accuracy: 2071/2280 (90.83%)\n",
      "Train Epoch 54: avg_discr_loss = -0.14483, avg_pred_loss = 0.194, avg_total_loss = 0.229\n",
      "Test set: Average loss: 0.2219429, Accuracy: 2092/2280 (91.75%)\n",
      "Train Epoch 55: avg_discr_loss = -0.14816, avg_pred_loss = 0.142, avg_total_loss = 0.151\n",
      "Test set: Average loss: 0.2218996, Accuracy: 2091/2280 (91.71%)\n",
      "Train Epoch 56: avg_discr_loss = -0.26935, avg_pred_loss = 0.151, avg_total_loss = 0.299\n",
      "Test set: Average loss: 0.2453881, Accuracy: 2100/2280 (92.11%)\n",
      "Train Epoch 57: avg_discr_loss = -0.17124, avg_pred_loss = 0.140, avg_total_loss = 0.203\n",
      "Test set: Average loss: 0.2216854, Accuracy: 2098/2280 (92.02%)\n",
      "Train Epoch 58: avg_discr_loss = -0.14660, avg_pred_loss = 0.136, avg_total_loss = 0.141\n",
      "Test set: Average loss: 0.2151560, Accuracy: 2107/2280 (92.41%)\n",
      "Train Epoch 59: avg_discr_loss = -0.20444, avg_pred_loss = 0.135, avg_total_loss = 0.206\n",
      "Test set: Average loss: 0.2379337, Accuracy: 2086/2280 (91.49%)\n",
      "Train Epoch 60: avg_discr_loss = -0.16430, avg_pred_loss = 0.144, avg_total_loss = 0.183\n",
      "Test set: Average loss: 0.2225178, Accuracy: 2098/2280 (92.02%)\n",
      "Train Epoch 61: avg_discr_loss = -0.13720, avg_pred_loss = 0.123, avg_total_loss = 0.120\n",
      "Test set: Average loss: 0.2343807, Accuracy: 2092/2280 (91.75%)\n",
      "Train Epoch 62: avg_discr_loss = -0.27518, avg_pred_loss = 0.144, avg_total_loss = 0.300\n",
      "Test set: Average loss: 0.2276060, Accuracy: 2111/2280 (92.59%)\n",
      "Train Epoch 63: avg_discr_loss = -0.15898, avg_pred_loss = 0.140, avg_total_loss = 0.194\n",
      "Test set: Average loss: 0.2493451, Accuracy: 2074/2280 (90.96%)\n",
      "Train Epoch 64: avg_discr_loss = -0.13209, avg_pred_loss = 0.131, avg_total_loss = 0.127\n",
      "Test set: Average loss: 0.2153635, Accuracy: 2109/2280 (92.50%)\n",
      "Train Epoch 65: avg_discr_loss = -0.16797, avg_pred_loss = 0.125, avg_total_loss = 0.157\n",
      "Test set: Average loss: 0.2555409, Accuracy: 2080/2280 (91.23%)\n",
      "Train Epoch 66: avg_discr_loss = -0.17281, avg_pred_loss = 0.125, avg_total_loss = 0.169\n",
      "Test set: Average loss: 0.2217841, Accuracy: 2111/2280 (92.59%)\n",
      "Train Epoch 67: avg_discr_loss = -0.13615, avg_pred_loss = 0.122, avg_total_loss = 0.120\n",
      "Test set: Average loss: 0.2196011, Accuracy: 2101/2280 (92.15%)\n",
      "Train Epoch 68: avg_discr_loss = -0.13915, avg_pred_loss = 0.115, avg_total_loss = 0.120\n",
      "Test set: Average loss: 0.2154305, Accuracy: 2116/2280 (92.81%)\n",
      "Train Epoch 69: avg_discr_loss = -0.16572, avg_pred_loss = 0.117, avg_total_loss = 0.150\n",
      "Test set: Average loss: 0.2143667, Accuracy: 2119/2280 (92.94%)\n",
      "Train Epoch 70: avg_discr_loss = -0.14857, avg_pred_loss = 0.107, avg_total_loss = 0.127\n",
      "Test set: Average loss: 0.2073409, Accuracy: 2130/2280 (93.42%)\n",
      "Train Epoch 71: avg_discr_loss = -0.12569, avg_pred_loss = 0.115, avg_total_loss = 0.110\n",
      "Test set: Average loss: 0.2055445, Accuracy: 2122/2280 (93.07%)\n",
      "Train Epoch 72: avg_discr_loss = -0.14609, avg_pred_loss = 0.118, avg_total_loss = 0.138\n",
      "Test set: Average loss: 0.2199815, Accuracy: 2109/2280 (92.50%)\n",
      "Train Epoch 73: avg_discr_loss = -0.15893, avg_pred_loss = 0.113, avg_total_loss = 0.148\n",
      "Test set: Average loss: 0.2024335, Accuracy: 2127/2280 (93.29%)\n",
      "Train Epoch 74: avg_discr_loss = -0.12152, avg_pred_loss = 0.110, avg_total_loss = 0.103\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set: Average loss: 0.1971523, Accuracy: 2131/2280 (93.46%)\n",
      "Train Epoch 75: avg_discr_loss = -0.13549, avg_pred_loss = 0.097, avg_total_loss = 0.105\n",
      "Test set: Average loss: 0.1936016, Accuracy: 2129/2280 (93.38%)\n",
      "Train Epoch 76: avg_discr_loss = -0.14823, avg_pred_loss = 0.101, avg_total_loss = 0.124\n",
      "Test set: Average loss: 0.2211935, Accuracy: 2119/2280 (92.94%)\n",
      "Train Epoch 77: avg_discr_loss = -0.14046, avg_pred_loss = 0.111, avg_total_loss = 0.127\n",
      "Test set: Average loss: 0.1979825, Accuracy: 2132/2280 (93.51%)\n",
      "Train Epoch 78: avg_discr_loss = -0.13091, avg_pred_loss = 0.102, avg_total_loss = 0.106\n",
      "Test set: Average loss: 0.1858655, Accuracy: 2135/2280 (93.64%)\n",
      "Train Epoch 79: avg_discr_loss = -0.12174, avg_pred_loss = 0.092, avg_total_loss = 0.092\n",
      "Test set: Average loss: 0.1981658, Accuracy: 2134/2280 (93.60%)\n",
      "Train Epoch 80: avg_discr_loss = -0.12566, avg_pred_loss = 0.111, avg_total_loss = 0.116\n",
      "Test set: Average loss: 0.1834375, Accuracy: 2133/2280 (93.55%)\n",
      "Train Epoch 81: avg_discr_loss = -0.12260, avg_pred_loss = 0.095, avg_total_loss = 0.104\n",
      "Test set: Average loss: 0.1866895, Accuracy: 2144/2280 (94.04%)\n",
      "Train Epoch 82: avg_discr_loss = -0.12953, avg_pred_loss = 0.101, avg_total_loss = 0.113\n",
      "Test set: Average loss: 0.2062894, Accuracy: 2108/2280 (92.46%)\n",
      "Train Epoch 83: avg_discr_loss = -0.12584, avg_pred_loss = 0.109, avg_total_loss = 0.117\n",
      "Test set: Average loss: 0.1919840, Accuracy: 2134/2280 (93.60%)\n",
      "Train Epoch 84: avg_discr_loss = -0.13018, avg_pred_loss = 0.108, avg_total_loss = 0.117\n",
      "Test set: Average loss: 0.1798227, Accuracy: 2143/2280 (93.99%)\n",
      "Train Epoch 85: avg_discr_loss = -0.12832, avg_pred_loss = 0.094, avg_total_loss = 0.101\n",
      "Test set: Average loss: 0.1925408, Accuracy: 2137/2280 (93.73%)\n",
      "Train Epoch 86: avg_discr_loss = -0.12216, avg_pred_loss = 0.099, avg_total_loss = 0.107\n",
      "Test set: Average loss: 0.1772581, Accuracy: 2146/2280 (94.12%)\n",
      "Train Epoch 87: avg_discr_loss = -0.12424, avg_pred_loss = 0.088, avg_total_loss = 0.096\n",
      "Test set: Average loss: 0.1878205, Accuracy: 2136/2280 (93.68%)\n",
      "Train Epoch 88: avg_discr_loss = -0.11376, avg_pred_loss = 0.086, avg_total_loss = 0.090\n",
      "Test set: Average loss: 0.1866092, Accuracy: 2139/2280 (93.82%)\n",
      "Train Epoch 89: avg_discr_loss = -0.10244, avg_pred_loss = 0.090, avg_total_loss = 0.087\n",
      "Test set: Average loss: 0.1759840, Accuracy: 2157/2280 (94.61%)\n",
      "Train Epoch 90: avg_discr_loss = -0.11166, avg_pred_loss = 0.084, avg_total_loss = 0.086\n",
      "Test set: Average loss: 0.1843264, Accuracy: 2143/2280 (93.99%)\n",
      "Train Epoch 91: avg_discr_loss = -0.10848, avg_pred_loss = 0.072, avg_total_loss = 0.072\n",
      "Test set: Average loss: 0.2097197, Accuracy: 2124/2280 (93.16%)\n",
      "Train Epoch 92: avg_discr_loss = -0.10539, avg_pred_loss = 0.080, avg_total_loss = 0.080\n",
      "Test set: Average loss: 0.1840039, Accuracy: 2146/2280 (94.12%)\n",
      "Train Epoch 93: avg_discr_loss = -0.09822, avg_pred_loss = 0.081, avg_total_loss = 0.079\n",
      "Test set: Average loss: 0.1709391, Accuracy: 2162/2280 (94.82%)\n",
      "Train Epoch 94: avg_discr_loss = -0.11730, avg_pred_loss = 0.086, avg_total_loss = 0.093\n",
      "Test set: Average loss: 0.1581234, Accuracy: 2177/2280 (95.48%)\n",
      "Train Epoch 95: avg_discr_loss = -0.11642, avg_pred_loss = 0.077, avg_total_loss = 0.079\n",
      "Test set: Average loss: 0.1739160, Accuracy: 2161/2280 (94.78%)\n",
      "Train Epoch 96: avg_discr_loss = -0.11694, avg_pred_loss = 0.072, avg_total_loss = 0.072\n",
      "Test set: Average loss: 0.1575189, Accuracy: 2183/2280 (95.75%)\n",
      "Train Epoch 97: avg_discr_loss = -0.11960, avg_pred_loss = 0.077, avg_total_loss = 0.080\n",
      "Test set: Average loss: 0.1916895, Accuracy: 2138/2280 (93.77%)\n",
      "Train Epoch 98: avg_discr_loss = -0.11993, avg_pred_loss = 0.073, avg_total_loss = 0.075\n",
      "Test set: Average loss: 0.1644848, Accuracy: 2170/2280 (95.18%)\n",
      "Train Epoch 99: avg_discr_loss = -0.11882, avg_pred_loss = 0.069, avg_total_loss = 0.072\n",
      "Test set: Average loss: 0.1701409, Accuracy: 2175/2280 (95.39%)\n",
      "Train Epoch 100: avg_discr_loss = -0.11930, avg_pred_loss = 0.068, avg_total_loss = 0.070\n",
      "Test set: Average loss: 0.1701062, Accuracy: 2166/2280 (95.00%)\n",
      "Test set: Average loss: 0.1701062, Accuracy: 2166/2280 (95.00%)\n"
     ]
    }
   ],
   "source": [
    "torch.autograd.set_detect_anomaly(True)\n",
    "\n",
    "# Set up optimizer\n",
    "opt_D = optim.Adam(discriminator.parameters(), lr = args.lr) # lr \n",
    "opt_non_D = optim.Adam(list(encoder.parameters()) + list(predictor.parameters()), lr = args.lr) # lr \n",
    "lr_scheduler_D = lr_scheduler.ExponentialLR(optimizer=opt_D, gamma=0.5 ** (1/(args.gamma_exp*(train_discr_step_extra+1)) * slow_lrD_decay))\n",
    "lr_scheduler_non_D = lr_scheduler.ExponentialLR(optimizer=opt_non_D, gamma=0.5 ** (1/args.gamma_exp))\n",
    "\n",
    "ind = list(range(args.batch_size))\n",
    "ind_test = list(range(1000))\n",
    "bce = nn.BCELoss()\n",
    "mse = nn.MSELoss()\n",
    "\n",
    "# Training loop\n",
    "def train(epoch):\n",
    "    for model in models:\n",
    "        model.train()\n",
    "    sum_discr_loss = 0\n",
    "    sum_total_loss = 0\n",
    "    sum_pred_loss = 0\n",
    "    for batch_idx, data_tuple in enumerate(train_loader):\n",
    "        if args.cuda:\n",
    "            data_tuple = tuple(ele.cuda() for ele in data_tuple)\n",
    "        if normalize:\n",
    "            data_raw, target, domain, data, mask = data_tuple\n",
    "        else:\n",
    "            data, target, domain, mask = data_tuple\n",
    "\n",
    "        # FF encoder and predictor\n",
    "        encoding = encoder((data, domain))\n",
    "        prediction = predictor((encoding, domain))\n",
    "\n",
    "        if use_label_noise:\n",
    "            noise = (torch.randn(domain.size()).cuda() * label_noise_std).unsqueeze(1)\n",
    "\n",
    "        # train discriminator\n",
    "        train_discr_step = 0\n",
    "        while args.dis_lambda > 0.0:\n",
    "            train_discr_step += 1\n",
    "            discr_pred_m, discr_pred_s = discriminator((encoding, domain))\n",
    "            discr_loss = gaussian_loss(discr_pred_m, discr_pred_s, domain.unsqueeze(1) / norm, np.mean(train_list) / norm, norm)\n",
    "            for model in models:\n",
    "                model.zero_grad()\n",
    "            discr_loss.backward(retain_graph=True)\n",
    "            opt_D.step()\n",
    "\n",
    "            # handle extra steps to train the discr's variance branch\n",
    "            if train_discr_step_extra > 0:\n",
    "                cur_extra_step = 0\n",
    "                while True:\n",
    "                    discr_pred_m, discr_pred_s = discriminator((encoding, domain))\n",
    "                    discr_loss = gaussian_loss(discr_pred_m.detach(), discr_pred_s, domain.unsqueeze(1) / norm)\n",
    "                    for model in models:\n",
    "                        model.zero_grad()\n",
    "                    discr_loss.backward(retain_graph=True)\n",
    "                    opt_D.step()\n",
    "                    cur_extra_step += 1\n",
    "                    if cur_extra_step > train_discr_step_extra:\n",
    "                        break\n",
    "\n",
    "            if discr_loss.item() < 1.1 * discr_thres and train_discr_step >= train_discr_step_tot:\n",
    "                sum_discr_loss += discr_loss.item()\n",
    "                break\n",
    "\n",
    "        # handle wgan\n",
    "        if args.wgan == 'wgan':\n",
    "            for p in discriminator.parameters():\n",
    "                p.data.clamp_(args.clamp_lower, args.clamp_upper)\n",
    "\n",
    "        # train encoder and predictor\n",
    "        pred_loss = masked_cross_entropy(prediction, target, mask)\n",
    "        discr_pred_m, discr_pred_s = discriminator((encoding, domain))\n",
    "        ent_loss = 0\n",
    "\n",
    "        discr_loss = gaussian_loss(discr_pred_m, discr_pred_s, domain.unsqueeze(1) / norm)\n",
    "        total_loss = pred_loss - discr_loss * args.dis_lambda\n",
    "\n",
    "        for model in models:\n",
    "            model.zero_grad()\n",
    "        total_loss.backward()\n",
    "        opt_non_D.step()\n",
    "        sum_pred_loss += pred_loss.item()\n",
    "        sum_total_loss += total_loss.item()\n",
    "\n",
    "    lr_scheduler_D.step()\n",
    "    lr_scheduler_non_D.step()\n",
    "\n",
    "    avg_discr_loss = sum_discr_loss / len(train_loader.dataset) * args.batch_size\n",
    "    avg_pred_loss = sum_pred_loss / len(train_loader.dataset) * args.batch_size\n",
    "    avg_total_loss = sum_total_loss / len(train_loader.dataset) * args.batch_size\n",
    "    log_txt = 'Train Epoch {}: avg_discr_loss = {:.5f}, avg_pred_loss = {:.3f}, avg_total_loss = {:.3f}'.format(epoch, avg_discr_loss, avg_pred_loss, avg_total_loss)\n",
    "    print(log_txt)\n",
    "    plain_log(args.log_file,log_txt+'\\n')\n",
    "    if epoch % args.save_interval == 0 and epoch != 0:\n",
    "        torch.save(encoder, '%s.model_enc' % args.save_head)\n",
    "        torch.save(predictor, '%s.model_pred' % args.save_head)\n",
    "        torch.save(discriminator, '%s.model_discr' % args.save_head)\n",
    "\n",
    "# Testing loop\n",
    "def test():\n",
    "    for model in models:\n",
    "        model.eval()\n",
    "    test_loss = 0\n",
    "    rmse_loss = 0\n",
    "    correct = 0\n",
    "    l_data = []\n",
    "    l_label = []\n",
    "    l_gt = []\n",
    "    l_encoding = []\n",
    "    l_domain = []\n",
    "    l_prob = []\n",
    "    #for data, target, domain in test_loader:\n",
    "    for data_tuple in test_loader:\n",
    "        if args.cuda:\n",
    "            data_tuple = tuple(ele.cuda() for ele in data_tuple)\n",
    "        if normalize:\n",
    "            data_raw, target, domain, data = data_tuple\n",
    "        else:\n",
    "            data, target, domain = data_tuple\n",
    "            data_raw = data\n",
    "        encoding = encoder((data, domain))\n",
    "        prediction = predictor((encoding, domain))\n",
    "        test_loss += F.nll_loss(prediction, target, reduction='sum').item() # sum up batch loss\n",
    "        pred = prediction.data.max(1, keepdim=True)[1] # get the index of the max log-probability\n",
    "        correct += pred.eq(target.view_as(pred)).cpu().sum()\n",
    "\n",
    "        l_data.append(data_raw.cpu().numpy())\n",
    "        l_label.append(pred.cpu().numpy())\n",
    "        l_gt.append(target.cpu().numpy())\n",
    "        l_encoding.append(encoding.data.cpu().numpy())\n",
    "        l_domain.append(domain.data.cpu().numpy())\n",
    "        l_prob.append(prediction.data.cpu().numpy())\n",
    "\n",
    "    data_all = np.concatenate(l_data, axis=0)\n",
    "    label_all = np.concatenate(l_label, axis=0)\n",
    "    gt_all = np.concatenate(l_gt, axis=0)\n",
    "    encoding_all = np.concatenate(l_encoding, axis=0)\n",
    "    domain_all = np.concatenate(l_domain, axis=0)\n",
    "    prob_all = np.concatenate(l_prob, axis=0)\n",
    "    d_pkl = dict()\n",
    "    d_pkl['data'] = data_all\n",
    "    d_pkl['label'] = label_all[:, 0]\n",
    "    d_pkl['gt'] = gt_all\n",
    "    d_pkl['encoding'] = encoding_all\n",
    "    d_pkl['domain'] = domain_all\n",
    "    d_pkl['prob'] = prob_all\n",
    "    write_pickle(d_pkl, fname_save)\n",
    "\n",
    "    test_loss /= len(test_loader.dataset)\n",
    "    log_txt = 'Test set: Average loss: {:.7f}, Accuracy: {}/{} ({:.2f}%)'.format(\n",
    "        test_loss, correct, len(test_loader.dataset),\n",
    "        100. * correct / len(test_loader.dataset))\n",
    "    print(log_txt)\n",
    "    plain_log(args.log_file,log_txt+'\\n')\n",
    "\n",
    "if args.checkpoint != 'none':\n",
    "    encoder = torch.load(args.checkpoint + '_enc')\n",
    "    predictor = torch.load(args.checkpoint + '_pred')\n",
    "    discriminator = torch.load(args.checkpoint + '_discr')\n",
    "    opt_D = optim.Adam(discriminator.parameters(), lr = args.lr) # lr \n",
    "    opt_non_D = optim.Adam(list(encoder.parameters()) + list(predictor.parameters()), lr = args.lr) # lr \n",
    "    lr_scheduler_D = lr_scheduler.ExponentialLR(optimizer=opt_D, gamma=0.5 ** (1/(args.gamma_exp*(train_discr_step_extra+1)) * slow_lrD_decay))\n",
    "    lr_scheduler_non_D = lr_scheduler.ExponentialLR(optimizer=opt_non_D, gamma=0.5 ** (1/args.gamma_exp))\n",
    "    models = [encoder, predictor, discriminator]\n",
    "    for model in models:\n",
    "        for key, module in model._modules.items():\n",
    "            print('key', key)\n",
    "            print('module', module)\n",
    "\n",
    "if not args.evaluate:\n",
    "    for epoch in range(1, args.epochs + 1):\n",
    "        train(epoch)\n",
    "        if epoch % 1 == 0:\n",
    "            test()\n",
    "test()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2280, 2)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x324 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize the model result\n",
    "from plot import plot_data_and_label\n",
    "info = read_pickle('pred_tmp.pkl')\n",
    "fig, ax = plt.subplots(1, 1, figsize=(6, 6 * 0.75))\n",
    "print(info['data'].shape)\n",
    "plot_data_and_label(ax, info['data'], info['label'])\n",
    "plt.title(f\"Model's prediction: CIDA\", fontsize=20)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}