first_play.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "### Imports"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "from tqdm import tqdm\n",
    "from torch.optim import Adam\n",
    "from torchvision.datasets import MNIST\n",
    "from torchvision.transforms import Compose, ToTensor, Normalize, Lambda\n",
    "from torch.utils.data import DataLoader\n",
    "import main\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "os.environ[\"KMP_DUPLICATE_LIB_OK\"]=\"TRUE\""
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Dataset loading"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\anaconda\\lib\\site-packages\\torchvision\\datasets\\mnist.py:498: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at  ..\\torch\\csrc\\utils\\tensor_numpy.cpp:180.)\n",
      "  return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([50000, 784])\n",
      "torch.Size([50000])\n"
     ]
    }
   ],
   "source": [
    "torch.manual_seed(1234)\n",
    "train_loader, test_loader = main.MNIST_loaders()\n",
    "x, y = next(iter(train_loader))\n",
    "print(x.shape)\n",
    "print(y.shape)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([50000, 784])\n",
      "torch.Size([50000, 784])\n"
     ]
    }
   ],
   "source": [
    "x_pos = main.overlay_y_on_x(x, y)\n",
    "print(x_pos.shape)\n",
    "rnd = torch.randperm(x.size(0))\n",
    "x_neg = main.overlay_y_on_x(x, y[rnd])\n",
    "print(x_neg.shape)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Display positive and negative samples"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.8214867\n",
      "-0.42421296\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 432x432 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x432 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_pos_sample = x_pos[4, :].reshape((28, 28)).numpy()\n",
    "print(x_pos_sample.max())\n",
    "print(x_pos_sample.min())\n",
    "x_neg_sample = x_neg[4, :].reshape((28, 28)).numpy()\n",
    "fig1 = plt.figure(0, figsize=(6, 6))\n",
    "plt.imshow(x_pos_sample)\n",
    "plt.show()\n",
    "fig2 = plt.figure(1, figsize=(6, 6))\n",
    "plt.imshow(x_neg_sample)\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}