ium_434766/lab5.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "orig_nbformat": 2,
  "kernelspec": {
   "name": "python385jvsc74a57bd02cef13873963874fd5439bd04a135498d1dd9725d9d90f40de0b76178a8e03b1",
   "display_name": "Python 3.8.5 64-bit (conda)"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import torch\n",
    "import torch.nn.functional as F\n",
    "from torch import nn\n",
    "from torch.autograd import Variable\n",
    "import torchvision.transforms as transforms\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import accuracy_score\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "\n",
    "\n",
    "class LogisticRegressionModel(nn.Module):\n",
    "    def __init__(self, input_dim, output_dim):\n",
    "        super(LogisticRegressionModel, self).__init__()\n",
    "        self.linear = nn.Linear(input_dim, output_dim)\n",
    "        self.sigmoid = nn.Sigmoid()\n",
    "    def forward(self, x):\n",
    "        out = self.linear(x)\n",
    "        return self.sigmoid(out)\n",
    "\n",
    "np.set_printoptions(suppress=False)\n",
    "data_train = pd.read_csv(\"data_train.csv\")\n",
    "data_test = pd.read_csv(\"data_test.csv\")\n",
    "data_val = pd.read_csv(\"data_val.csv\")\n",
    "FEATURES =  [ 'age','hypertension','heart_disease','ever_married', 'avg_glucose_level', 'bmi']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = data_train[FEATURES].astype(np.float32)\n",
    "y_train = data_train['stroke'].astype(np.float32)\n",
    "\n",
    "x_test = data_test[FEATURES].astype(np.float32)\n",
    "y_test = data_test['stroke'].astype(np.float32)\n",
    "\n",
    "\n",
    "\n",
    "fTrain = torch.from_numpy(x_train.values)\n",
    "tTrain = torch.from_numpy(y_train.values.reshape(2945,1))\n",
    "\n",
    "fTest= torch.from_numpy(x_test.values)\n",
    "tTest = torch.from_numpy(y_test.values)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "batch_size = 150\n",
    "n_iters = 1000\n",
    "num_epochs = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_dim = 6\n",
    "output_dim = 1\n",
    "\n",
    "model = LogisticRegressionModel(input_dim, output_dim)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "learning_rate = 0.001\n",
    "\n",
    "criterion = torch.nn.BCELoss(reduction='mean') \n",
    "optimizer = torch.optim.SGD(model.parameters(), lr = learning_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "torch.Size([1, 6])\ntorch.Size([1])\n"
     ]
    }
   ],
   "source": [
    "print(list(model.parameters())[0].size())\n",
    "print(list(model.parameters())[1].size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch # 0\n0.34391772747039795\nEpoch # 1\n0.3400452435016632\nEpoch # 2\n0.33628249168395996\nEpoch # 3\n0.3326331079006195\nEpoch # 4\n0.3291005790233612\nEpoch # 5\n0.32568827271461487\nEpoch # 6\n0.32239940762519836\nEpoch # 7\n0.3192369043827057\nEpoch # 8\n0.3162035048007965\nEpoch # 9\n0.31330153346061707\n"
     ]
    }
   ],
   "source": [
    "for epoch in range(num_epochs):\n",
    "    print (\"Epoch #\",epoch)\n",
    "    model.train()\n",
    "    optimizer.zero_grad()\n",
    "    # Forward pass\n",
    "    y_pred = model(fTrain)\n",
    "    # Compute Loss\n",
    "    loss = criterion(y_pred, tTrain)\n",
    "    print(loss.item())\n",
    "    # Backward pass\n",
    "    loss.backward()\n",
    "    optimizer.step()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "predicted Y value:  tensor([[0.0089],\n        [0.0051],\n        [0.1535],\n        [0.1008],\n        [0.0365],\n        [0.0014],\n        [0.1275],\n        [0.0172],\n        [0.1439],\n        [0.0088],\n        [0.0013],\n        [0.3466],\n        [0.0078],\n        [0.0303],\n        [0.0024],\n        [0.0607],\n        [0.0556],\n        [0.0826],\n        [0.0765],\n        [0.0027],\n        [0.0869],\n        [0.0424],\n        [0.0013],\n        [0.1338],\n        [0.0017],\n        [0.0020],\n        [0.0009],\n        [0.0014],\n        [0.0090],\n        [0.4073],\n        [0.0026],\n        [0.0009],\n        [0.0141],\n        [0.0897],\n        [0.3593],\n        [0.3849],\n        [0.0073],\n        [0.0204],\n        [0.1406],\n        [0.0053],\n        [0.3840],\n        [0.0802],\n        [0.0068],\n        [0.0190],\n        [0.3849],\n        [0.0034],\n        [0.0045],\n        [0.3272],\n        [0.0397],\n        [0.3087],\n        [0.0162],\n        [0.0159],\n        [0.0033],\n        [0.0559],\n        [0.0238],\n        [0.0073],\n        [0.0113],\n        [0.0102],\n        [0.3827],\n        [0.0359],\n        [0.0138],\n        [0.0248],\n        [0.0080],\n        [0.1858],\n        [0.0766],\n        [0.0123],\n        [0.0077],\n        [0.0042],\n        [0.0908],\n        [0.4172],\n        [0.0010],\n        [0.1105],\n        [0.0463],\n        [0.1457],\n        [0.0078],\n        [0.0821],\n        [0.0011],\n        [0.0210],\n        [0.0273],\n        [0.0248],\n        [0.0082],\n        [0.0007],\n        [0.0022],\n        [0.2436],\n        [0.0297],\n        [0.0235],\n        [0.0168],\n        [0.0053],\n        [0.0128],\n        [0.0156],\n        [0.0009],\n        [0.0375],\n        [0.0008],\n        [0.0645],\n        [0.0750],\n        [0.0055],\n        [0.0185],\n        [0.0008],\n        [0.0082],\n        [0.0138],\n        [0.2082],\n        [0.1823],\n        [0.0027],\n        [0.0124],\n        [0.0010],\n        [0.0187],\n        [0.2454],\n        [0.0019],\n        [0.1413],\n        [0.0010],\n        [0.0050],\n        [0.0020],\n        [0.0011],\n        [0.2266],\n        [0.0545],\n        [0.0164],\n        [0.0678],\n        [0.0012],\n        [0.0271],\n        [0.0009],\n        [0.0029],\n        [0.0058],\n        [0.0009],\n        [0.0762],\n        [0.0013],\n        [0.0276],\n        [0.3940],\n        [0.4213],\n        [0.0041],\n        [0.0144],\n        [0.1491],\n        [0.0011],\n        [0.0077],\n        [0.0032],\n        [0.3155],\n        [0.0009],\n        [0.0072],\n        [0.0056],\n        [0.3580],\n        [0.3235],\n        [0.0130],\n        [0.4032],\n        [0.0405],\n        [0.2882],\n        [0.0045],\n        [0.0041],\n        [0.0026],\n        [0.0354],\n        [0.0094],\n        [0.0278],\n        [0.0011],\n        [0.0036],\n        [0.2996],\n        [0.0652],\n        [0.4247],\n        [0.0048],\n        [0.0016],\n        [0.0703],\n        [0.3676],\n        [0.0231],\n        [0.0206],\n        [0.0093],\n        [0.0087],\n        [0.0649],\n        [0.0207],\n        [0.0010],\n        [0.0076],\n        [0.3366],\n        [0.0015],\n        [0.0034],\n        [0.2819],\n        [0.0007],\n        [0.0024],\n        [0.0015],\n        [0.1623],\n        [0.0838],\n        [0.0431],\n        [0.2744],\n        [0.1369],\n        [0.0007],\n        [0.0022],\n        [0.2049],\n        [0.0010],\n        [0.1057],\n        [0.0503],\n        [0.0021],\n        [0.0136],\n        [0.1939],\n        [0.0401],\n        [0.0010],\n        [0.4003],\n        [0.2621],\n        [0.0087],\n        [0.3507],\n        [0.0061],\n        [0.0012],\n        [0.0103],\n        [0.0080],\n        [0.1068],\n        [0.0098],\n        [0.2625],\n        [0.0162],\n        [0.0178],\n        [0.0215],\n        [0.0283],\n        [0.0444],\n        [0.0356],\n        [0.0037],\n        [0.0779],\n        [0.0652],\n        [0.0521],\n        [0.3626],\n        [0.0116],\n        [0.2099],\n  
     ]
    }
   ],
   "source": [
    "\n",
    "y_pred = model(fTest)\n",
    "print(\"predicted Y value: \", y_pred.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "The accuracy is 0.9480651731160896\n"
     ]
    }
   ],
   "source": [
    "print (\"The accuracy is\", accuracy_score(tTest, np.argmax(y_pred.detach().numpy(), axis=1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(model, 'stroke.pkl')"
   ]
  }
 ]
}