lstm-autoencoder-train.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import warnings\n",
    "import math\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from denoising.autoencoder import *\n",
    "from denoising.dataset import SeqWindowDataset\n",
    "\n",
    "from utils.metrics import get_metrics\n",
    "from data_prep.yahoo_fin_api import load_df\n",
    "\n",
    "%matplotlib inline\n",
    "warnings.filterwarnings(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(                  Open        High         Low       Close   Adj Close  \\\n",
       " Date                                                                     \n",
       " 1990-02-26  324.160004  328.670013  323.980011  328.670013  328.670013   \n",
       " 1990-02-27  328.679993  331.940002  328.470001  330.260010  330.260010   \n",
       " 1990-02-28  330.260010  333.480011  330.160004  331.890015  331.890015   \n",
       " 1990-03-01  331.890015  334.399994  331.079987  332.739990  332.739990   \n",
       " 1990-03-02  332.739990  335.540009  332.720001  335.540009  335.540009   \n",
       " \n",
       "                Volume        Mean  \n",
       " Date                               \n",
       " 1990-02-26  148900000  326.370010  \n",
       " 1990-02-27  152590000  329.837502  \n",
       " 1990-02-28  184400000  331.447510  \n",
       " 1990-03-01  157930000  332.527496  \n",
       " 1990-03-02  164330000  334.135002  ,\n",
       " 8313)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = load_df('data/snp.csv')\n",
    "df.head(), len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Date\n",
      "2013-03-25    1554.927490\n",
      "2013-03-26    1557.774963\n",
      "2013-03-27    1560.642487\n",
      "2013-03-28    1565.852478\n",
      "2013-04-01    1565.097504\n",
      "Name: Mean, dtype: float64\n",
      "==========\n",
      "Date\n",
      "2013-04-02    1567.062531\n",
      "2013-04-03    1561.302490\n",
      "2013-04-04    1557.197479\n",
      "2013-04-05    1553.184998\n",
      "2013-04-08    1557.007477\n",
      "Name: Mean, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "TRAIN_SIZE = 0.7\n",
    "cut_idx = df.iloc[int(len(df) * TRAIN_SIZE)].name\n",
    "data = df['Mean']\n",
    "data_min, data_max = data.min(), data.max()\n",
    "#train_data = data[data.index <= cut_idx]\n",
    "#test_data = data[cut_idx < data.index]\n",
    "\n",
    "#print(train_data.tail())\n",
    "#print('=' * 10)\n",
    "#print(test_data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 326.3700],\n",
       "        [ 329.8375],\n",
       "        [ 331.4475],\n",
       "        ...,\n",
       "        [4024.3076],\n",
       "        [3996.7876],\n",
       "        [4007.1025]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = torch.tensor(df['Mean'], dtype=torch.float32).view(-1, 1)\n",
    "#train_data = torch.tensor(train_data, dtype=torch.float32).view(-1, 1)\n",
    "#test_data = torch.tensor(test_data, dtype=torch.float32).view(-1, 1)\n",
    "\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7fda0af6fdf0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "data = data.to(device)\n",
    "torch.manual_seed(0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LSTM Stacked AutoEncoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "WINDOW_SIZE = 30\n",
    "\n",
    "num_epoch = 450"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "lstm_encoder = LstmEncoder(WINDOW_SIZE).to(device)\n",
    "lstm_decoder = LstmDecoder(WINDOW_SIZE).to(device)\n",
    "\n",
    "criterion = nn.MSELoss()\n",
    "optimizer = optim.Adam(list(lstm_encoder.parameters()) + list(lstm_decoder.parameters()), lr=1e-5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LSTM Stacked AutoEncoder\n",
      "LstmEncoder(\n",
      "  (rnn1): LSTM(1, 128, num_layers=5, batch_first=True)\n",
      "  (rnn2): LSTM(128, 64, num_layers=5, batch_first=True)\n",
      ")\n",
      "==========================================================================================\n",
      "LstmDecoder(\n",
      "  (rnn1): LSTM(1, 1, num_layers=5, batch_first=True)\n",
      "  (rnn2): LSTM(1, 2, num_layers=5, batch_first=True)\n",
      "  (output_layer): Linear(in_features=2, out_features=1, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "print('LSTM Stacked AutoEncoder')\n",
    "print(lstm_encoder)\n",
    "print('=' * 90)\n",
    "print(lstm_decoder)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "ename": "RuntimeError",
     "evalue": "shape '[32, 64]' is invalid for input of size 10240",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[12], line 11\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[39mfor\u001b[39;00m seq \u001b[39min\u001b[39;00m dataloader:\n\u001b[1;32m     10\u001b[0m     seq \u001b[39m=\u001b[39m seq\u001b[39m.\u001b[39mto(device)\n\u001b[0;32m---> 11\u001b[0m     output \u001b[39m=\u001b[39m lstm_encoder(seq)\n\u001b[1;32m     12\u001b[0m     pred \u001b[39m=\u001b[39m lstm_decoder(output)\n\u001b[1;32m     14\u001b[0m     loss \u001b[39m=\u001b[39m criterion(pred, data)\n",
      "File \u001b[0;32m~/.local/lib/python3.10/site-packages/torch/nn/modules/module.py:1190\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1186\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1187\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1188\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1189\u001b[0m         \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1190\u001b[0m     \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39;49m\u001b[39minput\u001b[39;49m, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m   1191\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1192\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n",
      "File \u001b[0;32m~/projects/Fin-Data-Denoising/denoising/autoencoder.py:87\u001b[0m, in \u001b[0;36mLstmEncoder.forward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m     84\u001b[0m x, (_, _) \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mrnn1(x)\n\u001b[1;32m     85\u001b[0m x, (hidden_n, _) \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mrnn2(x)\n\u001b[0;32m---> 87\u001b[0m \u001b[39mreturn\u001b[39;00m hidden_n\u001b[39m.\u001b[39;49mview(batch_size, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49membedding_dim)\n",
      "\u001b[0;31mRuntimeError\u001b[0m: shape '[32, 64]' is invalid for input of size 10240"
     ]
    }
   ],
   "source": [
    "train_loss = []\n",
    "dataset = SeqWindowDataset(data, WINDOW_SIZE)\n",
    "dataloader = DataLoader(dataset, batch_size=32, shuffle=False)\n",
    "lstm_encoder.train()\n",
    "lstm_decoder.train()\n",
    "\n",
    "plot_epoch = num_epoch - np.floor(num_epoch * np.log10(np.linspace(1, 10, 20)))\n",
    "for epoch in range(num_epoch):\n",
    "    for seq in dataloader:\n",
    "        seq = seq.to(device)\n",
    "        output = lstm_encoder(seq)\n",
    "        pred = lstm_decoder(output)\n",
    "\n",
    "        loss = criterion(pred, data)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        train_loss.append(loss.detach().cpu().numpy())\n",
    "\n",
    "        if epoch in plot_epoch or epoch == num_epoch - 1:\n",
    "            df['LSTMPredMean'] = (data).detach().cpu().numpy().squeeze()\n",
    "            plt.ylim([data_min, data_max])\n",
    "            df[['Mean', 'LSTMPredMean']].plot(color=['b', 'r'], figsize=(18, 10))\n",
    "            plt.savefig(f'plots/LSTM-kernel51-epoch{epoch:03}.png')\n",
    "\n",
    "train_loss = pd.Series(train_loss).astype(np.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_ = train_loss.plot.line()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "1990-02-26    0.001783\n",
       "1990-02-27    0.000366\n",
       "1990-02-28    0.000126\n",
       "1990-03-01    0.000072\n",
       "1990-03-02    0.000053\n",
       "                ...   \n",
       "2023-02-16    0.000002\n",
       "2023-02-17    0.000002\n",
       "2023-02-21    0.000002\n",
       "2023-02-22    0.000002\n",
       "2023-02-23    0.000002\n",
       "Name: LSTMPredMean, Length: 8313, dtype: float32"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['LSTMPredMean']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['LSTMPredMean'] = lstm_autoencoder(data).detach().cpu().numpy().squeeze()\n",
    "#df['LSTMPredMean'] = lstm_decoder(lstm_encoder(data)).detach().cpu().numpy().squeeze()\n",
    "_ = df[['Mean', 'LSTMPredMean']].plot(color=['b', 'r'], figsize=(18, 10))"
   ]
  }
 ],
 "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": "3.10.6"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}