2_train_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "CONFIG_FILE = \"model/model_v_0_5_config.py\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'model.model_v_0_5_config' from '/Users/michaelbonon/Coding/stock-price-predictor/apps/stock-price-model/model/model_v_0_5_config.py'>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import importlib\n",
    "import numpy as np\n",
    "import json\n",
    "import os\n",
    "from typing import Any\n",
    "from src.helpers import load_metadata, load_preprocessed_data, merge_set, standardize\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "config = importlib.import_module(CONFIG_FILE.replace(\"/\", \".\").replace(\".py\", \"\"))\n",
    "importlib.reload(config)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['BASELINE_MODEL_FILENAME', 'DATA_DIR', 'DATA_FILENAME_MAP', 'GET_INDEX_X_Y', 'LEARNING_RATE', 'LOAD_BASELINE_MODEL', 'LOSS_FUNCTION', 'MAX_EPOCHS', 'METRICS', 'MODELS_DIR', 'Metadata', 'PATIENCE', 'TEST_SET', 'TRAINING_SET', 'TRAINING_SET_PREPROCESSING', 'VALIDATION_SET', 'VERSION_MAJOR', 'VERSION_MINOR', '__builtins__', '__cached__', '__doc__', '__file__', '__loader__', '__name__', '__package__', '__spec__', 'df_windowed_to_index_x_y', 'pd']\n"
     ]
    }
   ],
   "source": [
    "# Check if config file has all the required attributes\n",
    "print(dir(config))\n",
    "\n",
    "required_attributes = [\n",
    "    \"LEARNING_RATE\",\n",
    "    \"MAX_EPOCHS\",\n",
    "    \"PATIENCE\",\n",
    "    \"VERSION_MAJOR\",\n",
    "    \"VERSION_MINOR\",\n",
    "    \"DATA_DIR\",\n",
    "    \"DATA_FILENAME_MAP\",\n",
    "    \"TRAINING_SET\",\n",
    "    \"TEST_SET\",\n",
    "    \"VALIDATION_SET\",\n",
    "    \"MODELS_DIR\",\n",
    "    \"LOAD_BASELINE_MODEL\",\n",
    "    \"GET_INDEX_X_Y\",\n",
    "    \"TRAINING_SET_PREPROCESSING\",\n",
    "    \"LOSS_FUNCTION\",\n",
    "    \"METRICS\",\n",
    "]\n",
    "\n",
    "for attr in required_attributes:\n",
    "    if not hasattr(config, attr):\n",
    "        raise ValueError(f\"Missing attribute {attr} in config file.\")\n",
    "    \n",
    "if config.LOAD_BASELINE_MODEL and not hasattr(config, \"BASELINE_MODEL_FILENAME\"):\n",
    "    raise ValueError(\"Missing attribute BASELINE_MODEL_FILENAME in config file.\")\n",
    "\n",
    "LEARNING_RATE = config.LEARNING_RATE\n",
    "MAX_EPOCHS = config.MAX_EPOCHS\n",
    "PATIENCE = config.PATIENCE\n",
    "VERSION_MAJOR = config.VERSION_MAJOR\n",
    "VERSION_MINOR = config.VERSION_MINOR\n",
    "DATA_DIR = config.DATA_DIR\n",
    "DATA_FILENAME_MAP = config.DATA_FILENAME_MAP\n",
    "TRAINING_SET = config.TRAINING_SET\n",
    "TEST_SET = config.TEST_SET\n",
    "VALIDATION_SET = config.VALIDATION_SET\n",
    "MODELS_DIR = config.MODELS_DIR\n",
    "LOAD_BASELINE_MODEL = config.LOAD_BASELINE_MODEL\n",
    "BASELINE_MODEL_FILENAME = config.BASELINE_MODEL_FILENAME if LOAD_BASELINE_MODEL else \"\"\n",
    "GET_INDEX_X_Y = config.GET_INDEX_X_Y\n",
    "TRAINING_SET_PREPROCESSING = config.TRAINING_SET_PREPROCESSING\n",
    "LOSS_FUNCTION = config.LOSS_FUNCTION\n",
    "METRICS = config.METRICS\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x290e13400>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "loaded_dataset = {}\n",
    "\n",
    "COMBINED_SETS = TRAINING_SET + VALIDATION_SET + TEST_SET\n",
    "\n",
    "for item in COMBINED_SETS:\n",
    "    filename = DATA_FILENAME_MAP[item][\"filename\"]\n",
    "    metadata_filename = DATA_FILENAME_MAP[item][\"metadata\"]\n",
    "\n",
    "    metadata = load_metadata(DATA_DIR, metadata_filename)\n",
    "    df = load_preprocessed_data(DATA_DIR, filename, metadata)\n",
    "    \n",
    "    index_data, feature_data, target_data = GET_INDEX_X_Y(df, metadata)\n",
    "\n",
    "    loaded_dataset[item] = {\n",
    "        \"index\": index_data,\n",
    "        \"x\": feature_data,\n",
    "        \"y\": target_data,\n",
    "    }\n",
    "\n",
    "    plt.plot(index_data, target_data)\n",
    "\n",
    "plt.legend(COMBINED_SETS)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = merge_set(TRAINING_SET, \"x\", loaded_dataset)\n",
    "train_y = merge_set(TRAINING_SET, \"y\", loaded_dataset)\n",
    "val_x = merge_set(VALIDATION_SET, \"x\", loaded_dataset)\n",
    "val_y = merge_set(VALIDATION_SET, \"y\", loaded_dataset)\n",
    "test_x = merge_set(TEST_SET, \"x\", loaded_dataset)\n",
    "test_y = merge_set(TEST_SET, \"y\", loaded_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "model_output {'Volume_mean': '2601463.2', 'Volume_std': '8855982.0'}\n"
     ]
    }
   ],
   "source": [
    "\n",
    "metadata_filename = DATA_FILENAME_MAP[TRAINING_SET[0]][\"metadata\"]\n",
    "metadata = load_metadata(DATA_DIR, metadata_filename)\n",
    "\n",
    "# Sample value of TRAINING_SET_PREPROCESSING\n",
    "# TRAINING_SET_PREPROCESSING = [\n",
    "#     (\"Volume\", [\"z-score\"]),\n",
    "# ]\n",
    "\n",
    "model_output = {}\n",
    "\n",
    "for feature_name, preprocessing_steps in TRAINING_SET_PREPROCESSING:\n",
    "    if feature_name not in metadata[\"features\"]:\n",
    "        raise ValueError(f\"Feature {feature_name} not found in metadata.\")\n",
    "    \n",
    "    target_index = metadata[\"features\"].index(feature_name)\n",
    "    for preprocessing_step in preprocessing_steps:\n",
    "        if preprocessing_step == \"z-score\":\n",
    "            training_mean = np.mean(train_x[:, :, target_index])\n",
    "            training_std = np.std(train_x[:, :, target_index])\n",
    "            standardize(train_x, target_index, training_mean, training_std)\n",
    "            standardize(val_x, target_index, training_mean, training_std)\n",
    "            model_output[f\"{feature_name}_mean\"] = str(training_mean)\n",
    "            model_output[f\"{feature_name}_std\"] = str(training_std)\n",
    "        else:\n",
    "            raise ValueError(f\"Preprocessing step {preprocessing_step} not supported.\")\n",
    "\n",
    "print(\"model_output\", model_output)\n",
    "\n",
    "model_output_params_filepath = \"{}/v_{}_{}/{}\".format(\n",
    "  MODELS_DIR,\n",
    "  VERSION_MAJOR,\n",
    "  VERSION_MINOR,\n",
    "  \"model_output_params.json\"\n",
    ")\n",
    "os.makedirs(os.path.dirname(model_output_params_filepath), exist_ok=True)\n",
    "with open(model_output_params_filepath, \"w\") as f:\n",
    "  json.dump(model_output, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-11 14:03:53.267133: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2023-03-11 14:03:53.416191: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   4/1822 [..............................] - ETA: 38s - loss: 79604.6094 - mean_absolute_error: 67.2525"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-11 14:03:53.553790: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1822/1822 [==============================] - ETA: 0s - loss: 15259.8555 - mean_absolute_error: 21.9295"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-11 14:04:15.006340: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2023-03-11 14:04:15.057652: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1822/1822 [==============================] - 30s 16ms/step - loss: 15259.8555 - mean_absolute_error: 21.9295 - val_loss: 5.7737 - val_mean_absolute_error: 0.9751\n",
      "Epoch 2/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 2089.9133 - mean_absolute_error: 9.0767 - val_loss: 3.4573 - val_mean_absolute_error: 0.7927\n",
      "Epoch 3/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 1400.9774 - mean_absolute_error: 8.4827 - val_loss: 2.4469 - val_mean_absolute_error: 0.6278\n",
      "Epoch 4/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 886.2618 - mean_absolute_error: 7.8272 - val_loss: 3.4616 - val_mean_absolute_error: 0.7406\n",
      "Epoch 5/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 754.2423 - mean_absolute_error: 7.5386 - val_loss: 2.6399 - val_mean_absolute_error: 0.6683\n",
      "Epoch 6/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 738.6957 - mean_absolute_error: 7.6448 - val_loss: 2.1007 - val_mean_absolute_error: 0.6018\n",
      "Epoch 7/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 808.1176 - mean_absolute_error: 8.1482 - val_loss: 1.8478 - val_mean_absolute_error: 0.5795\n",
      "Epoch 8/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 781.2180 - mean_absolute_error: 7.8018 - val_loss: 1.7564 - val_mean_absolute_error: 0.5704\n",
      "Epoch 9/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 623.9061 - mean_absolute_error: 6.7671 - val_loss: 3.2412 - val_mean_absolute_error: 0.7713\n",
      "Epoch 10/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 694.3646 - mean_absolute_error: 7.3365 - val_loss: 3.4421 - val_mean_absolute_error: 0.7156\n",
      "Epoch 11/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 586.5646 - mean_absolute_error: 6.9060 - val_loss: 3.8105 - val_mean_absolute_error: 0.7954\n",
      "Epoch 12/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 892.7903 - mean_absolute_error: 8.2097 - val_loss: 1.4628 - val_mean_absolute_error: 0.5054\n",
      "Epoch 13/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 771.1492 - mean_absolute_error: 7.6546 - val_loss: 2.2565 - val_mean_absolute_error: 0.5838\n",
      "Epoch 14/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 589.2891 - mean_absolute_error: 7.1039 - val_loss: 1.6706 - val_mean_absolute_error: 0.5452\n",
      "Epoch 15/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 732.0844 - mean_absolute_error: 7.5427 - val_loss: 1.6716 - val_mean_absolute_error: 0.5337\n",
      "Epoch 16/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 705.6791 - mean_absolute_error: 7.0302 - val_loss: 3.6708 - val_mean_absolute_error: 0.7333\n",
      "Epoch 17/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 934.6361 - mean_absolute_error: 8.4734 - val_loss: 1.7679 - val_mean_absolute_error: 0.5354\n",
      "Epoch 18/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 795.5216 - mean_absolute_error: 7.8582 - val_loss: 1.4938 - val_mean_absolute_error: 0.5019\n",
      "Epoch 19/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 801.8546 - mean_absolute_error: 7.8611 - val_loss: 2.7738 - val_mean_absolute_error: 0.6651\n",
      "Epoch 20/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 784.8078 - mean_absolute_error: 8.2087 - val_loss: 4.8239 - val_mean_absolute_error: 0.7953\n",
      "Epoch 21/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 792.1699 - mean_absolute_error: 7.6659 - val_loss: 1.7499 - val_mean_absolute_error: 0.5314\n",
      "Epoch 22/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 1081.9513 - mean_absolute_error: 7.8590 - val_loss: 3.3939 - val_mean_absolute_error: 0.7041\n",
      "Epoch 23/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 695.6683 - mean_absolute_error: 7.2676 - val_loss: 1.4761 - val_mean_absolute_error: 0.5382\n",
      "Epoch 24/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 571.7967 - mean_absolute_error: 7.1764 - val_loss: 1.5938 - val_mean_absolute_error: 0.5089\n",
      "Epoch 25/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 616.2241 - mean_absolute_error: 7.1695 - val_loss: 2.1046 - val_mean_absolute_error: 0.7093\n",
      "Epoch 26/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 539.2078 - mean_absolute_error: 6.9423 - val_loss: 2.2363 - val_mean_absolute_error: 0.6737\n",
      "Epoch 27/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 548.1458 - mean_absolute_error: 7.1482 - val_loss: 1.6290 - val_mean_absolute_error: 0.5795\n",
      "Epoch 28/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 401.7221 - mean_absolute_error: 6.3324 - val_loss: 1.9269 - val_mean_absolute_error: 0.6094\n",
      "Epoch 29/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 564.9283 - mean_absolute_error: 6.7390 - val_loss: 1.5403 - val_mean_absolute_error: 0.5333\n",
      "Epoch 30/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 531.8109 - mean_absolute_error: 6.7540 - val_loss: 8.5104 - val_mean_absolute_error: 1.1585\n",
      "Epoch 31/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 546.3727 - mean_absolute_error: 6.9044 - val_loss: 7.8805 - val_mean_absolute_error: 0.9844\n",
      "Epoch 32/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 463.5978 - mean_absolute_error: 6.4518 - val_loss: 1.5698 - val_mean_absolute_error: 0.5073\n",
      "Epoch 33/50\n",
      "1822/1822 [==============================] - 28s 15ms/step - loss: 998.6803 - mean_absolute_error: 7.7539 - val_loss: 3.2350 - val_mean_absolute_error: 0.7445\n",
      "Epoch 34/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 721.4807 - mean_absolute_error: 7.6725 - val_loss: 5.5103 - val_mean_absolute_error: 0.8508\n",
      "Epoch 35/50\n",
      "1822/1822 [==============================] - 28s 15ms/step - loss: 665.5878 - mean_absolute_error: 7.2417 - val_loss: 4.2470 - val_mean_absolute_error: 0.5150\n",
      "Epoch 36/50\n",
      "1822/1822 [==============================] - 28s 15ms/step - loss: 742.0114 - mean_absolute_error: 7.5628 - val_loss: 4.8855 - val_mean_absolute_error: 0.7071\n",
      "Epoch 37/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 628.6645 - mean_absolute_error: 7.1430 - val_loss: 6.0695 - val_mean_absolute_error: 0.5481\n",
      "Epoch 38/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 607.0881 - mean_absolute_error: 6.8731 - val_loss: 10.7904 - val_mean_absolute_error: 0.9966\n",
      "Epoch 39/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 722.5032 - mean_absolute_error: 7.3828 - val_loss: 39.6470 - val_mean_absolute_error: 0.6635\n",
      "Epoch 40/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 789.8147 - mean_absolute_error: 7.3204 - val_loss: 99.9308 - val_mean_absolute_error: 0.6947\n",
      "Epoch 41/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 594.4186 - mean_absolute_error: 7.0872 - val_loss: 118.3770 - val_mean_absolute_error: 1.1182\n",
      "Epoch 42/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 602.1065 - mean_absolute_error: 6.9114 - val_loss: 65.7768 - val_mean_absolute_error: 0.6982\n",
      "Epoch 43/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 599.5961 - mean_absolute_error: 7.0049 - val_loss: 59.9523 - val_mean_absolute_error: 0.8296\n",
      "Epoch 44/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 723.0894 - mean_absolute_error: 7.6687 - val_loss: 29.3982 - val_mean_absolute_error: 1.0019\n",
      "Epoch 45/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 615.0386 - mean_absolute_error: 6.9301 - val_loss: 26.9372 - val_mean_absolute_error: 0.6431\n",
      "Epoch 46/50\n",
      "1822/1822 [==============================] - 60s 33ms/step - loss: 599.5033 - mean_absolute_error: 7.2173 - val_loss: 30.2030 - val_mean_absolute_error: 0.7945\n",
      "Epoch 47/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 654.8567 - mean_absolute_error: 7.4103 - val_loss: 18.2509 - val_mean_absolute_error: 0.7652\n",
      "Epoch 48/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 665.4899 - mean_absolute_error: 7.1017 - val_loss: 24.2787 - val_mean_absolute_error: 0.7229\n",
      "Epoch 49/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 724.8076 - mean_absolute_error: 7.4743 - val_loss: 7.8869 - val_mean_absolute_error: 0.7617\n",
      "Epoch 50/50\n",
      "1822/1822 [==============================] - 27s 15ms/step - loss: 578.1404 - mean_absolute_error: 6.6631 - val_loss: 8.5377 - val_mean_absolute_error: 0.6482\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2903db700>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from keras import layers, losses, metrics, callbacks, models, optimizers\n",
    "\n",
    "feature_size = len(metadata[\"features\"])\n",
    "window_size = metadata[\"window_size\"]\n",
    "\n",
    "model = None\n",
    "\n",
    "if LOAD_BASELINE_MODEL:\n",
    "    model = models.load_model(f\"{MODELS_DIR}/{BASELINE_MODEL_FILENAME}\")\n",
    "else:\n",
    "    model = models.Sequential(\n",
    "        [\n",
    "            layers.Input(shape=(window_size, feature_size)),\n",
    "            layers.LSTM(128),\n",
    "            layers.Dense(64, activation=\"relu\"),\n",
    "            layers.Dense(32, activation=\"relu\"),\n",
    "            layers.Dense(1),\n",
    "        ]\n",
    "    )\n",
    "\n",
    "def compile_and_fit(\n",
    "    model: models.Sequential,\n",
    "    train_x: np.ndarray[str, Any],\n",
    "    train_y: np.ndarray[str, Any],\n",
    "    val_x: np.ndarray[str, Any],\n",
    "    val_y: np.ndarray[str, Any],\n",
    "    max_epochs: int,\n",
    "    patience: int,\n",
    "    learning_rate: float,\n",
    "):\n",
    "    early_stopping = callbacks.EarlyStopping(\n",
    "        monitor=\"val_loss\", patience=patience, mode=\"min\", restore_best_weights=True\n",
    "    )\n",
    "\n",
    "    model.compile(\n",
    "        loss=LOSS_FUNCTION,\n",
    "        optimizer=optimizers.Adam(learning_rate=learning_rate),\n",
    "        metrics=METRICS,\n",
    "    )\n",
    "\n",
    "    history = model.fit(\n",
    "        train_x,\n",
    "        train_y,\n",
    "        epochs=max_epochs,\n",
    "        validation_data=(val_x, val_y),\n",
    "        callbacks=[early_stopping],\n",
    "    )\n",
    "    return history\n",
    "\n",
    "\n",
    "compile_and_fit(model, train_x, train_y, val_x, val_y, MAX_EPOCHS, PATIENCE, LEARNING_RATE)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   1/1589 [..............................] - ETA: 9:30"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-11 14:27:11.493446: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2023-03-11 14:27:11.564373: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1589/1589 [==============================] - 6s 4ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "val_predictions = None\n",
    "\n",
    "if model is not None:\n",
    "    val_predictions = model.predict(val_x)\n",
    "    plt.plot(val_predictions)\n",
    "    plt.plot(val_y)\n",
    "    plt.legend(['Validation Predictions', 'Validation Observations'])\n",
    "    plt.show()\n",
    "else:\n",
    "    print(\"No model to evaluate\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm_1 (LSTM)               (None, 128)               68608     \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 64)                8256      \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 32)                2080      \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 78,977\n",
      "Trainable params: 78,977\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model version: 0.5\n",
      "Learning rate:  0.0001\n",
      "Max epochs:  50\n",
      "Patience:  50\n",
      "Training set:  ['FB', 'MAXS', 'AC', 'SMPH', 'BDO', 'GLO', 'DITO', 'PAL', 'MER', 'SCC', 'SEVN', 'GMA7']\n",
      "Validation set:  ['URC', 'SMC', 'MEG', 'MBT', 'TBGI', '2GO', 'FGEN', 'PX', 'RRHI', 'MB']\n",
      "Load baseline model:  True\n",
      "Performance on validation set\n",
      "['loss', 'mean_absolute_error']\n",
      "[8.538, 0.648]\n"
     ]
    }
   ],
   "source": [
    "model.summary()\n",
    "\n",
    "print(\"Model version: {}.{}\".format(VERSION_MAJOR, VERSION_MINOR))\n",
    "print(\"Learning rate: \", LEARNING_RATE)\n",
    "print(\"Max epochs: \", MAX_EPOCHS)\n",
    "print(\"Patience: \", PATIENCE)\n",
    "print(\"Training set: \", TRAINING_SET)\n",
    "print(\"Validation set: \", VALIDATION_SET)\n",
    "print(\"Load baseline model: \", LOAD_BASELINE_MODEL)\n",
    "\n",
    "results = model.evaluate(val_x, val_y, verbose=0)\n",
    "# Round the results to 3 decimal places.\n",
    "results = [round(result, 3) for result in results]\n",
    "print(\"Performance on validation set\")\n",
    "print(model.metrics_names)\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save the trained model into a file so we can use it to make predictions later\n",
    "model_filepath = \"{}/v_{}_{}/{}\".format(\n",
    "  MODELS_DIR,\n",
    "  VERSION_MAJOR,\n",
    "  VERSION_MINOR,\n",
    "  \"model.h5\"\n",
    ")\n",
    "os.makedirs(os.path.dirname(model_filepath), exist_ok=True)\n",
    "model.save(\n",
    "  model_filepath,\n",
    "  save_format=\"h5\",\n",
    ")\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "stock-price-model",
   "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": "3.10.7"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "c9651fe4711116b58576197cb7b9a712ffc15ff381142ea8887467b64161274a"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}