7.1-graph-neural-network-forecasting-stock.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "from torch_geometric.data import Data\n",
    "import pandas as pd\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch_geometric.nn import GCNConv"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.299368Z",
     "start_time": "2024-07-18T06:17:55.743596Z"
    }
   },
   "id": "7681350084ff31a8",
   "execution_count": 1
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "initial_id",
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.306754Z",
     "start_time": "2024-07-18T06:17:57.300410Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['AAPL', 'AMZN', 'GOOGL', 'META', 'NFLX'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "# Load your data\n",
    "data = pd.read_csv('data/stocks1.csv', index_col='Date', parse_dates=True)\n",
    "# Compute daily returns\n",
    "returns = data.pct_change().dropna()\n",
    "print(returns.columns)\n",
    "# Let's assume S&P index is in the first column\n",
    "target_series = returns['AMZN']\n",
    "features = returns.drop(columns=['AAPL', 'AMZN', 'GOOGL', 'META', 'NFLX'])"
   ]
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "#!pip install torch_geometric"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.308424Z",
     "start_time": "2024-07-18T06:17:57.306672Z"
    }
   },
   "id": "646abd2a9e600b2c",
   "execution_count": 3
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([], size=(2, 0), dtype=torch.int64)\n"
     ]
    }
   ],
   "source": [
    "# Compute correlation matrix\n",
    "correlation_matrix = features.corr().values\n",
    "\n",
    "# Create graph edges based on correlation\n",
    "edge_index = np.array(np.nonzero(correlation_matrix > 0.5)).T  # Using 0.5 as a threshold\n",
    "edge_index = torch.tensor(edge_index, dtype=torch.long).t().contiguous()\n",
    "print(edge_index)\n",
    "# Node features\n",
    "x = torch.tensor(features.values, dtype=torch.float)\n",
    "\n",
    "# Target: future values of the S&P index\n",
    "y = torch.tensor(target_series.shift(-1).dropna().values, dtype=torch.float).unsqueeze(1)\n",
    "\n",
    "# Update node features to match target length\n",
    "x = x[:-1]  # Exclude the last row to match target length\n",
    "\n",
    "# Define the graph data object\n",
    "graph_data = Data(x=x, edge_index=edge_index)\n"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.312933Z",
     "start_time": "2024-07-18T06:17:57.310226Z"
    }
   },
   "id": "88937c68de0a37e4",
   "execution_count": 4
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "class GCN(nn.Module):\n",
    "    def __init__(self, in_channels, hidden_channels, out_channels):\n",
    "        super(GCN, self).__init__()\n",
    "        self.conv1 = GCNConv(in_channels, hidden_channels)\n",
    "        self.conv2 = GCNConv(hidden_channels, out_channels)\n",
    "\n",
    "    def forward(self, x, edge_index):\n",
    "        x = self.conv1(x, edge_index)\n",
    "        x = F.relu(x)\n",
    "        x = self.conv2(x, edge_index)\n",
    "        return x\n",
    "\n",
    "# Initialize model, loss, and optimizer\n",
    "model = GCN(in_channels=graph_data.num_features, hidden_channels=16, out_channels=1)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=0.01, weight_decay=5e-4)\n",
    "criterion = nn.MSELoss()\n"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.335713Z",
     "start_time": "2024-07-18T06:17:57.314773Z"
    }
   },
   "id": "1b2f10a80f63a4c1",
   "execution_count": 5
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0, Loss: 0.0005640679737553\n",
      "Epoch 10, Loss: 0.0005829180008731782\n",
      "Epoch 20, Loss: 0.0005707201780751348\n",
      "Epoch 30, Loss: 0.0005645995843224227\n",
      "Epoch 40, Loss: 0.0005636165151372552\n",
      "Epoch 50, Loss: 0.000563537934795022\n",
      "Epoch 60, Loss: 0.0005635461420752108\n",
      "Epoch 70, Loss: 0.00056353589752689\n",
      "Epoch 80, Loss: 0.0005634989938698709\n",
      "Epoch 90, Loss: 0.0005634866538457572\n",
      "Epoch 100, Loss: 0.000563488865736872\n",
      "Epoch 110, Loss: 0.0005634863628074527\n",
      "Epoch 120, Loss: 0.0005634866538457572\n",
      "Epoch 130, Loss: 0.0005634864210151136\n",
      "Epoch 140, Loss: 0.0005634863628074527\n",
      "Epoch 150, Loss: 0.0005634864210151136\n",
      "Epoch 160, Loss: 0.0005634864210151136\n",
      "Epoch 170, Loss: 0.0005634864210151136\n",
      "Epoch 180, Loss: 0.0005634863628074527\n",
      "Epoch 190, Loss: 0.0005634863628074527\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 1200x600 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.train()\n",
    "num_epochs = 200\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    optimizer.zero_grad()\n",
    "    out = model(graph_data.x, graph_data.edge_index)\n",
    "    loss = criterion(out, y)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    if epoch % 10 == 0:\n",
    "        print(f'Epoch {epoch}, Loss: {loss.item()}')\n",
    "        \n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    predictions = model(graph_data.x, graph_data.edge_index).squeeze().numpy()\n",
    "\n",
    "# Compare predictions with actual values\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "actual_values = y.squeeze().numpy()\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.plot(actual_values[0:20], label='Actual S&P')\n",
    "plt.plot(predictions[0:20], label='Predicted S&P')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.668787Z",
     "start_time": "2024-07-18T06:17:57.335442Z"
    }
   },
   "id": "6a3bb1c639a3b4b",
   "execution_count": 6
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "# This work needs serious rework (possibly hyperparameter tuning) ++"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.670664Z",
     "start_time": "2024-07-18T06:17:57.669166Z"
    }
   },
   "id": "67e979fa42f2bb6e",
   "execution_count": 7
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "# Draw graph network \n"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-07-18T06:17:57.675069Z",
     "start_time": "2024-07-18T06:17:57.671697Z"
    }
   },
   "id": "3ddb3798e7fd21cd",
   "execution_count": 8
  }
 ],
 "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": 5
}