modelling_and_visualization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "from torch.autograd import Variable\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import StandardScaler   \n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on the GPU\n"
     ]
    }
   ],
   "source": [
    "if torch.cuda.is_available():\n",
    "    device = torch.device(\"cuda:0\") \n",
    "    print(\"Running on the GPU\")\n",
    "else:\n",
    "    device = torch.device(\"cpu\")\n",
    "    print(\"Running on the CPU\")\n",
    "# device = torch.device(\"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['x', 'y', 'val'], dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('batman_logo_dataset.csv').sample(frac=1).astype(int)\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df[['x', 'y']].values\n",
    "Y = df[['val']].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n",
      "<class 'numpy.ndarray'>\n",
      "(72900, 2)\n",
      "(72900, 1)\n"
     ]
    }
   ],
   "source": [
    "print(type(X))\n",
    "print(type(Y))\n",
    "print(X.shape)\n",
    "print(Y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x203733cf128>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X[:,0].reshape(X.shape[0], 1), X[:,1].reshape(X.shape[0], 1), c=Y, s=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "X = scaler.fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([72900, 2])\n",
      "torch.Size([72900, 1])\n"
     ]
    }
   ],
   "source": [
    "X = torch.Tensor(X)\n",
    "Y = torch.Tensor(Y)\n",
    "print(X.shape)\n",
    "print(Y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Net(nn.Module):\n",
    "    \n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.inp = nn.Linear(2, 256)\n",
    "        self.mid = nn.Linear(256, 256)\n",
    "        self.out = nn.Linear(256, 1)\n",
    "    \n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.inp(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = F.relu(self.mid(x))\n",
    "        x = self.out(x)\n",
    "        return x\n",
    "    \n",
    "model = Net()\n",
    "model.to(device)\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.001)     \n",
    "EPOCHS = 100\n",
    "batch_size = 512\n",
    "criterion = nn.BCEWithLogitsLoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def binary_acc(y_pred, y_test):\n",
    "    y_pred_tag = torch.round(torch.sigmoid(y_pred))\n",
    "\n",
    "    correct_results_sum = (y_pred_tag == y_test).sum().float()\n",
    "    acc = correct_results_sum/y_test.shape[0]\n",
    "    acc = acc * 100\n",
    "    \n",
    "    return acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(X, Y, model, EPOCHS, batch_size, criterion, optimizer):\n",
    "    losses = []\n",
    "    accuracies = []\n",
    "    for epoch in range(EPOCHS):\n",
    "        for i in range(0, X.shape[0], batch_size):\n",
    "            x , y = X[i:i+batch_size, :], Y[i:i+batch_size, :]\n",
    "            model.zero_grad()\n",
    "            optimizer.zero_grad()\n",
    "            y_hat = model.forward(x.to(device))\n",
    "            loss = criterion(y_hat,y.to(device))\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            if i%(batch_size*10) == 0:\n",
    "                losses.append(loss)\n",
    "                accuracies.append(binary_acc(y_hat, y.to(device)))\n",
    "    return model, accuracies, losses\n",
    "\n",
    "model, accuracies, losses = train(X, Y, model, EPOCHS, batch_size, criterion, optimizer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x203038f9e48>]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def predict(x):\n",
    "    #Convert into numpy element to tensor\n",
    "    x = torch.Tensor(x).to(device)\n",
    "    #Predict and return ans\n",
    "    ans = 1/(1+np.exp(-model.forward(x).cpu().detach().numpy()))\n",
    "    return ans\n",
    "plt.plot(losses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x20303972780>]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(accuracies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot():\n",
    "    \n",
    "    Y = predict(X)\n",
    "    Y = np.round(Y)\n",
    "    plt.scatter(X[:,0].detach().numpy().reshape(X.shape[0], 1), X[:,1].detach().numpy().reshape(X.shape[0], 1), c=Y, s=1)\n",
    "plot()"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}