src/models/ETH.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Users\\codya\\AppData\\Local\\Programs\\Python\\Python311\\Lib\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from tensorflow.keras.models import Sequential, Model\n",
    "from tensorflow.keras.layers import LSTM, Dense, Input\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import random\n",
    "\n",
    "import os \n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "122751\n"
     ]
    }
   ],
   "source": [
    "CURRENCY = \"ETH\"\n",
    "CLOSE = f\"Close_{CURRENCY}\"\n",
    "\n",
    "seed = random.randint(0, 1000000)\n",
    "seed = 122751\n",
    "print(seed)\n",
    "tf.random.set_seed(seed)\n",
    "np.random.seed(seed)\n",
    "random.seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_pickle = '../../data/federalfinancegestion/dataset.pkl'\n",
    "df = pd.read_pickle(dataset_pickle)\n",
    "\n",
    "# Remove all string (object type) columns from the dataset\n",
    "data_numeric = df.select_dtypes(exclude=['object'])\n",
    "\n",
    "# Finding the column that closely matches \"Close_BTC\" in the numeric dataset\n",
    "close_btc_column_numeric = [col for col in data_numeric.columns if \"Close_BTC\" in col]\n",
    "\n",
    "close_btc_column_numeric = close_btc_column_numeric[0]\n",
    "# Calculate the correlation of all columns with Close_BTC in the numeric dataset\n",
    "correlation_with_close_btc_numeric = data_numeric.corr()[close_btc_column_numeric].sort_values(ascending=False)\n",
    "\n",
    "# Extract the top 10 most correlated feature names\n",
    "top_feature = correlation_with_close_btc_numeric.head(10).index.tolist()\n",
    "include = top_feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data\n",
    "df.fillna(0, inplace=True)\n",
    "df['date'] = pd.to_datetime(df['date'])\n",
    "df.set_index('date', inplace=True)\n",
    "\n",
    "start_date_training = pd.to_datetime('2017-08-01')\n",
    "end_date_training = pd.to_datetime('2022-08-31')\n",
    "start_date_testing = pd.to_datetime('2022-09-01')\n",
    "end_date_testing = pd.to_datetime('2023-04-30')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = df.loc[start_date_training:end_date_training]\n",
    "test_df = df.loc[start_date_testing:end_date_testing]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculating the logarithmic difference of 'Close'\n",
    "df['Close_Log'] = np.log(df[CLOSE])  # First, compute the log of 'Close'\n",
    "df['Log_Diff_Close'] = df['Close_Log'].diff().fillna(0)  # Then compute the difference and handle NaN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Include features\n",
    "train = train_df[include].values\n",
    "test = test_df[include].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize the data\n",
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "train_scaled = scaler.fit_transform(train)\n",
    "test_scaled = scaler.transform(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Users\\codya\\AppData\\Local\\Programs\\Python\\Python311\\Lib\\site-packages\\keras\\src\\backend.py:1398: The name tf.executing_eagerly_outside_functions is deprecated. Please use tf.compat.v1.executing_eagerly_outside_functions instead.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Users\\codya\\AppData\\Local\\Programs\\Python\\Python311\\Lib\\site-packages\\keras\\src\\optimizers\\__init__.py:309: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
      "\n",
      "Epoch 1/150\n",
      "WARNING:tensorflow:From c:\\Users\\codya\\AppData\\Local\\Programs\\Python\\Python311\\Lib\\site-packages\\keras\\src\\utils\\tf_utils.py:492: The name tf.ragged.RaggedTensorValue is deprecated. Please use tf.compat.v1.ragged.RaggedTensorValue instead.\n",
      "\n",
      "33/33 [==============================] - 1s 7ms/step - loss: 0.1212 - val_loss: 0.0905\n",
      "Epoch 2/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0340 - val_loss: 0.0543\n",
      "Epoch 3/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0104 - val_loss: 0.0210\n",
      "Epoch 4/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0070 - val_loss: 0.0159\n",
      "Epoch 5/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0052 - val_loss: 0.0147\n",
      "Epoch 6/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0041 - val_loss: 0.0142\n",
      "Epoch 7/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0036 - val_loss: 0.0136\n",
      "Epoch 8/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0141\n",
      "Epoch 9/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0137\n",
      "Epoch 10/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0024 - val_loss: 0.0132\n",
      "Epoch 11/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0021 - val_loss: 0.0130\n",
      "Epoch 12/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0019 - val_loss: 0.0130\n",
      "Epoch 13/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0016 - val_loss: 0.0126\n",
      "Epoch 14/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0014 - val_loss: 0.0119\n",
      "Epoch 15/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0013 - val_loss: 0.0119\n",
      "Epoch 16/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0012 - val_loss: 0.0123\n",
      "Epoch 17/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0012 - val_loss: 0.0110\n",
      "Epoch 18/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0011 - val_loss: 0.0111\n",
      "Epoch 19/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0010 - val_loss: 0.0108\n",
      "Epoch 20/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 0.0010 - val_loss: 0.0120\n",
      "Epoch 21/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 9.6384e-04 - val_loss: 0.0108\n",
      "Epoch 22/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 9.1985e-04 - val_loss: 0.0111\n",
      "Epoch 23/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 8.9167e-04 - val_loss: 0.0112\n",
      "Epoch 24/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 8.7055e-04 - val_loss: 0.0108\n",
      "Epoch 25/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 8.2774e-04 - val_loss: 0.0106\n",
      "Epoch 26/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 8.0012e-04 - val_loss: 0.0107\n",
      "Epoch 27/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 8.1073e-04 - val_loss: 0.0098\n",
      "Epoch 28/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 7.5846e-04 - val_loss: 0.0106\n",
      "Epoch 29/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 7.0618e-04 - val_loss: 0.0102\n",
      "Epoch 30/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 6.8114e-04 - val_loss: 0.0100\n",
      "Epoch 31/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 6.4546e-04 - val_loss: 0.0102\n",
      "Epoch 32/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 6.2936e-04 - val_loss: 0.0106\n",
      "Epoch 33/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 5.8980e-04 - val_loss: 0.0097\n",
      "Epoch 34/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 5.5331e-04 - val_loss: 0.0092\n",
      "Epoch 35/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 5.2970e-04 - val_loss: 0.0095\n",
      "Epoch 36/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 5.1025e-04 - val_loss: 0.0108\n",
      "Epoch 37/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 4.8364e-04 - val_loss: 0.0103\n",
      "Epoch 38/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 4.8407e-04 - val_loss: 0.0091\n",
      "Epoch 39/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 4.5632e-04 - val_loss: 0.0100\n",
      "Epoch 40/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 4.2619e-04 - val_loss: 0.0093\n",
      "Epoch 41/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 4.1461e-04 - val_loss: 0.0086\n",
      "Epoch 42/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 4.2562e-04 - val_loss: 0.0102\n",
      "Epoch 43/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 4.1660e-04 - val_loss: 0.0095\n",
      "Epoch 44/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.8619e-04 - val_loss: 0.0087\n",
      "Epoch 45/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.8474e-04 - val_loss: 0.0090\n",
      "Epoch 46/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.7827e-04 - val_loss: 0.0087\n",
      "Epoch 47/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.7673e-04 - val_loss: 0.0088\n",
      "Epoch 48/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.6345e-04 - val_loss: 0.0085\n",
      "Epoch 49/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.5277e-04 - val_loss: 0.0086\n",
      "Epoch 50/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.4255e-04 - val_loss: 0.0086\n",
      "Epoch 51/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.4050e-04 - val_loss: 0.0083\n",
      "Epoch 52/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.3805e-04 - val_loss: 0.0098\n",
      "Epoch 53/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.2869e-04 - val_loss: 0.0084\n",
      "Epoch 54/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.2230e-04 - val_loss: 0.0086\n",
      "Epoch 55/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.2286e-04 - val_loss: 0.0086\n",
      "Epoch 56/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.1315e-04 - val_loss: 0.0084\n",
      "Epoch 57/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.1112e-04 - val_loss: 0.0077\n",
      "Epoch 58/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.0427e-04 - val_loss: 0.0081\n",
      "Epoch 59/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.0827e-04 - val_loss: 0.0079\n",
      "Epoch 60/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.9911e-04 - val_loss: 0.0084\n",
      "Epoch 61/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 3.0220e-04 - val_loss: 0.0087\n",
      "Epoch 62/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.9110e-04 - val_loss: 0.0082\n",
      "Epoch 63/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.8072e-04 - val_loss: 0.0082\n",
      "Epoch 64/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.8888e-04 - val_loss: 0.0084\n",
      "Epoch 65/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.7579e-04 - val_loss: 0.0085\n",
      "Epoch 66/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.7551e-04 - val_loss: 0.0085\n",
      "Epoch 67/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.6786e-04 - val_loss: 0.0082\n",
      "Epoch 68/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.6386e-04 - val_loss: 0.0082\n",
      "Epoch 69/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5932e-04 - val_loss: 0.0079\n",
      "Epoch 70/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.6639e-04 - val_loss: 0.0084\n",
      "Epoch 71/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5879e-04 - val_loss: 0.0085\n",
      "Epoch 72/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.6112e-04 - val_loss: 0.0088\n",
      "Epoch 73/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5270e-04 - val_loss: 0.0080\n",
      "Epoch 74/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5175e-04 - val_loss: 0.0082\n",
      "Epoch 75/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 2.5215e-04 - val_loss: 0.0090\n",
      "Epoch 76/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5939e-04 - val_loss: 0.0080\n",
      "Epoch 77/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.5012e-04 - val_loss: 0.0075\n",
      "Epoch 78/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.4430e-04 - val_loss: 0.0085\n",
      "Epoch 79/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3969e-04 - val_loss: 0.0081\n",
      "Epoch 80/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3803e-04 - val_loss: 0.0086\n",
      "Epoch 81/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.4452e-04 - val_loss: 0.0085\n",
      "Epoch 82/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3858e-04 - val_loss: 0.0080\n",
      "Epoch 83/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3465e-04 - val_loss: 0.0087\n",
      "Epoch 84/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3594e-04 - val_loss: 0.0082\n",
      "Epoch 85/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3345e-04 - val_loss: 0.0086\n",
      "Epoch 86/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.2919e-04 - val_loss: 0.0078\n",
      "Epoch 87/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.2609e-04 - val_loss: 0.0085\n",
      "Epoch 88/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3068e-04 - val_loss: 0.0080\n",
      "Epoch 89/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.4051e-04 - val_loss: 0.0083\n",
      "Epoch 90/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 2.3802e-04 - val_loss: 0.0079\n",
      "Epoch 91/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.1573e-04 - val_loss: 0.0071\n",
      "Epoch 92/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.2069e-04 - val_loss: 0.0081\n",
      "Epoch 93/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3374e-04 - val_loss: 0.0075\n",
      "Epoch 94/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3273e-04 - val_loss: 0.0083\n",
      "Epoch 95/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.2208e-04 - val_loss: 0.0084\n",
      "Epoch 96/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.1641e-04 - val_loss: 0.0082\n",
      "Epoch 97/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.0491e-04 - val_loss: 0.0086\n",
      "Epoch 98/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.1172e-04 - val_loss: 0.0073\n",
      "Epoch 99/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.3521e-04 - val_loss: 0.0082\n",
      "Epoch 100/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.1615e-04 - val_loss: 0.0083\n",
      "Epoch 101/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9903e-04 - val_loss: 0.0081\n",
      "Epoch 102/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9563e-04 - val_loss: 0.0086\n",
      "Epoch 103/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9869e-04 - val_loss: 0.0083\n",
      "Epoch 104/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9985e-04 - val_loss: 0.0087\n",
      "Epoch 105/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.0477e-04 - val_loss: 0.0080\n",
      "Epoch 106/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8950e-04 - val_loss: 0.0085\n",
      "Epoch 107/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.0299e-04 - val_loss: 0.0081\n",
      "Epoch 108/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 1.9421e-04 - val_loss: 0.0083\n",
      "Epoch 109/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9307e-04 - val_loss: 0.0082\n",
      "Epoch 110/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8547e-04 - val_loss: 0.0079\n",
      "Epoch 111/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8383e-04 - val_loss: 0.0080\n",
      "Epoch 112/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8572e-04 - val_loss: 0.0081\n",
      "Epoch 113/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8544e-04 - val_loss: 0.0083\n",
      "Epoch 114/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9539e-04 - val_loss: 0.0087\n",
      "Epoch 115/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.9066e-04 - val_loss: 0.0083\n",
      "Epoch 116/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8638e-04 - val_loss: 0.0083\n",
      "Epoch 117/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8432e-04 - val_loss: 0.0087\n",
      "Epoch 118/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8386e-04 - val_loss: 0.0081\n",
      "Epoch 119/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7805e-04 - val_loss: 0.0084\n",
      "Epoch 120/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7766e-04 - val_loss: 0.0079\n",
      "Epoch 121/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8361e-04 - val_loss: 0.0080\n",
      "Epoch 122/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8779e-04 - val_loss: 0.0084\n",
      "Epoch 123/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 2.1702e-04 - val_loss: 0.0078\n",
      "Epoch 124/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7587e-04 - val_loss: 0.0083\n",
      "Epoch 125/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7296e-04 - val_loss: 0.0082\n",
      "Epoch 126/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8058e-04 - val_loss: 0.0088\n",
      "Epoch 127/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.8986e-04 - val_loss: 0.0081\n",
      "Epoch 128/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7270e-04 - val_loss: 0.0077\n",
      "Epoch 129/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7527e-04 - val_loss: 0.0082\n",
      "Epoch 130/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6962e-04 - val_loss: 0.0081\n",
      "Epoch 131/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6518e-04 - val_loss: 0.0087\n",
      "Epoch 132/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6368e-04 - val_loss: 0.0084\n",
      "Epoch 133/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6554e-04 - val_loss: 0.0078\n",
      "Epoch 134/150\n",
      "33/33 [==============================] - 0s 3ms/step - loss: 1.8424e-04 - val_loss: 0.0083\n",
      "Epoch 135/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.7982e-04 - val_loss: 0.0076\n",
      "Epoch 136/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6047e-04 - val_loss: 0.0083\n",
      "Epoch 137/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6087e-04 - val_loss: 0.0080\n",
      "Epoch 138/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5765e-04 - val_loss: 0.0078\n",
      "Epoch 139/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5807e-04 - val_loss: 0.0084\n",
      "Epoch 140/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.6452e-04 - val_loss: 0.0081\n",
      "Epoch 141/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5774e-04 - val_loss: 0.0085\n",
      "Epoch 142/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5817e-04 - val_loss: 0.0078\n",
      "Epoch 143/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5458e-04 - val_loss: 0.0085\n",
      "Epoch 144/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5827e-04 - val_loss: 0.0079\n",
      "Epoch 145/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.4894e-04 - val_loss: 0.0083\n",
      "Epoch 146/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5159e-04 - val_loss: 0.0083\n",
      "Epoch 147/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.4477e-04 - val_loss: 0.0078\n",
      "Epoch 148/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.5186e-04 - val_loss: 0.0080\n",
      "Epoch 149/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.4384e-04 - val_loss: 0.0079\n",
      "Epoch 150/150\n",
      "33/33 [==============================] - 0s 2ms/step - loss: 1.4936e-04 - val_loss: 0.0086\n",
      "42/42 [==============================] - 0s 1ms/step\n"
     ]
    }
   ],
   "source": [
    "# # # SDAE\n",
    "input_layer = Input(shape=(len(include),))\n",
    "encoded = Dense(64, activation='relu')(input_layer)\n",
    "encoded = Dense(32, activation='relu')(encoded)\n",
    "encoded = Dense(16, activation='relu')(encoded)\n",
    "\n",
    "decoded = Dense(32, activation='relu')(encoded)\n",
    "decoded = Dense(64, activation='relu')(decoded)\n",
    "decoded = Dense(len(include), activation='sigmoid')(decoded)\n",
    "\n",
    "autoencoder = Model(input_layer, decoded)\n",
    "autoencoder.compile(optimizer='adam', loss='mse')\n",
    "\n",
    "# Train the autoencoder\n",
    "autoencoder.fit(train_scaled, train_scaled, epochs=150, batch_size=32, shuffle=True, validation_split=0.2)\n",
    "\n",
    "# Get the encoded representation of the data\n",
    "encoded_features = autoencoder.predict(train_scaled)\n",
    "\n",
    "def create_sequences(data, sequence_length):\n",
    "    xs, ys = [], []\n",
    "    for i in range(len(data) - sequence_length):\n",
    "        x = data[i:(i + sequence_length), :]\n",
    "        y = data[i + sequence_length, 0]\n",
    "        xs.append(x)\n",
    "        ys.append(y)\n",
    "    return np.array(xs), np.array(ys)\n",
    "\n",
    "sequence_length = 5  # Number of days used to predict the next day\n",
    "X_train, y_train = create_sequences(train_scaled, sequence_length)\n",
    "X_test, y_test = create_sequences(test_scaled, sequence_length)\n",
    "\n",
    "# Reshape inputs for LSTM [samples, time steps, features]\n",
    "X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], len(include)))\n",
    "X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], len(include)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm (LSTM)                 (None, 5, 50)             12200     \n",
      "                                                                 \n",
      " lstm_1 (LSTM)               (None, 50)                20200     \n",
      "                                                                 \n",
      " dense_6 (Dense)             (None, 25)                1275      \n",
      "                                                                 \n",
      " dense_7 (Dense)             (None, 1)                 26        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 33701 (131.64 KB)\n",
      "Trainable params: 33701 (131.64 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Build the LSTM model\n",
    "model = Sequential()\n",
    "model.add(LSTM(50, return_sequences=True, input_shape=(sequence_length, len(include))))\n",
    "model.add(LSTM(50, return_sequences=False))\n",
    "model.add(Dense(25))\n",
    "model.add(Dense(1))\n",
    "\n",
    "model.compile(optimizer='adam', loss='mean_squared_error')\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/150\n",
      "21/21 [==============================] - 3s 4ms/step - loss: 0.0241\n",
      "Epoch 2/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0052\n",
      "Epoch 3/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0033\n",
      "Epoch 4/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 5/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 6/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0012\n",
      "Epoch 7/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0012\n",
      "Epoch 8/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 9/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 9.7843e-04\n",
      "Epoch 10/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 11/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 12/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 0.0012\n",
      "Epoch 13/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0012\n",
      "Epoch 14/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 9.1198e-04\n",
      "Epoch 15/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.4839e-04\n",
      "Epoch 16/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 9.2613e-04\n",
      "Epoch 17/150\n",
      "21/21 [==============================] - 0s 6ms/step - loss: 9.1904e-04\n",
      "Epoch 18/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.6802e-04\n",
      "Epoch 19/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.7747e-04\n",
      "Epoch 20/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 8.9561e-04\n",
      "Epoch 21/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 7.9815e-04\n",
      "Epoch 22/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.4893e-04\n",
      "Epoch 23/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.7126e-04\n",
      "Epoch 24/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.0795e-04\n",
      "Epoch 25/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.1364e-04\n",
      "Epoch 26/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.5758e-04\n",
      "Epoch 27/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.6973e-04\n",
      "Epoch 28/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.3505e-04\n",
      "Epoch 29/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 9.1592e-04\n",
      "Epoch 30/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.7170e-04\n",
      "Epoch 31/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.4929e-04\n",
      "Epoch 32/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.4938e-04\n",
      "Epoch 33/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 0.0011\n",
      "Epoch 34/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.5463e-04\n",
      "Epoch 35/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.0288e-04\n",
      "Epoch 36/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.0372e-04\n",
      "Epoch 37/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.9809e-04\n",
      "Epoch 38/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.7464e-04\n",
      "Epoch 39/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.8837e-04\n",
      "Epoch 40/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 8.3061e-04\n",
      "Epoch 41/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.7300e-04\n",
      "Epoch 42/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 7.5791e-04\n",
      "Epoch 43/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.9651e-04\n",
      "Epoch 44/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 6.1366e-04\n",
      "Epoch 45/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 6.2372e-04\n",
      "Epoch 46/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.6276e-04\n",
      "Epoch 47/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.0428e-04\n",
      "Epoch 48/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.6010e-04\n",
      "Epoch 49/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.2767e-04\n",
      "Epoch 50/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.1235e-04\n",
      "Epoch 51/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.4930e-04\n",
      "Epoch 52/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.8182e-04\n",
      "Epoch 53/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 6.2473e-04\n",
      "Epoch 54/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.5020e-04\n",
      "Epoch 55/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.5492e-04\n",
      "Epoch 56/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.1511e-04\n",
      "Epoch 57/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.6874e-04\n",
      "Epoch 58/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.9887e-04\n",
      "Epoch 59/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 5.5956e-04\n",
      "Epoch 60/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.0439e-04\n",
      "Epoch 61/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.1668e-04\n",
      "Epoch 62/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.7773e-04\n",
      "Epoch 63/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.4751e-04\n",
      "Epoch 64/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.5360e-04\n",
      "Epoch 65/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.2889e-04\n",
      "Epoch 66/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.0007e-04\n",
      "Epoch 67/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.3804e-04\n",
      "Epoch 68/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.3421e-04\n",
      "Epoch 69/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.1371e-04\n",
      "Epoch 70/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.7432e-04\n",
      "Epoch 71/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.4371e-04\n",
      "Epoch 72/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.5146e-04\n",
      "Epoch 73/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.9319e-04\n",
      "Epoch 74/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.8278e-04\n",
      "Epoch 75/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 6.4715e-04\n",
      "Epoch 76/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.0730e-04\n",
      "Epoch 77/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.2227e-04\n",
      "Epoch 78/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.8937e-04\n",
      "Epoch 79/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.7733e-04\n",
      "Epoch 80/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.9739e-04\n",
      "Epoch 81/150\n",
      "21/21 [==============================] - 0s 5ms/step - loss: 4.6656e-04\n",
      "Epoch 82/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.0858e-04\n",
      "Epoch 83/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.6799e-04\n",
      "Epoch 84/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6761e-04\n",
      "Epoch 85/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.8408e-04\n",
      "Epoch 86/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6713e-04\n",
      "Epoch 87/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.8302e-04\n",
      "Epoch 88/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.3011e-04\n",
      "Epoch 89/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 7.3241e-04\n",
      "Epoch 90/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.7290e-04\n",
      "Epoch 91/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6788e-04\n",
      "Epoch 92/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6951e-04\n",
      "Epoch 93/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.5231e-04\n",
      "Epoch 94/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4272e-04\n",
      "Epoch 95/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.7132e-04\n",
      "Epoch 96/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.2907e-04\n",
      "Epoch 97/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.1806e-04\n",
      "Epoch 98/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.5062e-04\n",
      "Epoch 99/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.3681e-04\n",
      "Epoch 100/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4267e-04\n",
      "Epoch 101/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4014e-04\n",
      "Epoch 102/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6211e-04\n",
      "Epoch 103/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.1546e-04\n",
      "Epoch 104/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.2437e-04\n",
      "Epoch 105/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4803e-04\n",
      "Epoch 106/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4814e-04\n",
      "Epoch 107/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.2992e-04\n",
      "Epoch 108/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.9948e-04\n",
      "Epoch 109/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.9843e-04\n",
      "Epoch 110/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.5193e-04\n",
      "Epoch 111/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.2568e-04\n",
      "Epoch 112/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.1013e-04\n",
      "Epoch 113/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0638e-04\n",
      "Epoch 114/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.2885e-04\n",
      "Epoch 115/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.2992e-04\n",
      "Epoch 116/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.1375e-04\n",
      "Epoch 117/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.5671e-04\n",
      "Epoch 118/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.5500e-04\n",
      "Epoch 119/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7892e-04\n",
      "Epoch 120/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8875e-04\n",
      "Epoch 121/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0655e-04\n",
      "Epoch 122/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0494e-04\n",
      "Epoch 123/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8700e-04\n",
      "Epoch 124/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0927e-04\n",
      "Epoch 125/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.1522e-04\n",
      "Epoch 126/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.5581e-04\n",
      "Epoch 127/150\n",
      "21/21 [==============================] - 0s 6ms/step - loss: 3.8933e-04\n",
      "Epoch 128/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.1587e-04\n",
      "Epoch 129/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.0976e-04\n",
      "Epoch 130/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.9776e-04\n",
      "Epoch 131/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.6730e-04\n",
      "Epoch 132/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8025e-04\n",
      "Epoch 133/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.5721e-04\n",
      "Epoch 134/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 4.9514e-04\n",
      "Epoch 135/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 5.2629e-04\n",
      "Epoch 136/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0786e-04\n",
      "Epoch 137/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7201e-04\n",
      "Epoch 138/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8920e-04\n",
      "Epoch 139/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.4502e-04\n",
      "Epoch 140/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7793e-04\n",
      "Epoch 141/150\n",
      "21/21 [==============================] - 0s 3ms/step - loss: 3.6709e-04\n",
      "Epoch 142/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 4.0145e-04\n",
      "Epoch 143/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7213e-04\n",
      "Epoch 144/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8674e-04\n",
      "Epoch 145/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.8142e-04\n",
      "Epoch 146/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7391e-04\n",
      "Epoch 147/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.5463e-04\n",
      "Epoch 148/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.5714e-04\n",
      "Epoch 149/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.6257e-04\n",
      "Epoch 150/150\n",
      "21/21 [==============================] - 0s 4ms/step - loss: 3.7201e-04\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.History at 0x2cf20633910>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, batch_size=64, epochs=150)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5/5 [==============================] - 1s 3ms/step\n"
     ]
    }
   ],
   "source": [
    "# Predict values\n",
    "predictions = model.predict(X_test)\n",
    "predictions = scaler.inverse_transform(np.concatenate((predictions, np.zeros((predictions.shape[0], len(include)-1))), axis=1))[:, 0]\n",
    "\n",
    "# Slice the actual prices to align with the predictions\n",
    "actual_prices_aligned = df.loc[start_date_testing:end_date_testing, CLOSE][-len(predictions):]\n",
    "predictions = pd.DataFrame(predictions, index=actual_prices_aligned.index, columns=['Predicted_Close'])\n",
    "\n",
    "# Convert actual_prices_aligned to a DataFrame\n",
    "actual_prices_aligned_df = pd.DataFrame(actual_prices_aligned, columns=[CLOSE])\n",
    "\n",
    "# Calculating the logarithm of predictions and then the logarithmic difference\n",
    "actual_prices_aligned_df['Predicted_Close'] = predictions\n",
    "actual_prices_aligned_df['Predicted_Close'] = np.log(actual_prices_aligned_df['Predicted_Close'])\n",
    "log_diff_predicted = actual_prices_aligned_df['Predicted_Close'].diff().fillna(0)\n",
    "log_diff_actual = pd.DataFrame(df[\"Log_Diff_Close\"], index=actual_prices_aligned.index, columns=['Log_Diff_Close'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Actual Log Diff Close</th>\n",
       "      <th>Predicted_Close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-09-08</th>\n",
       "      <td>0.003289</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-09-09</th>\n",
       "      <td>0.049426</td>\n",
       "      <td>0.000614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-09-12</th>\n",
       "      <td>-0.001083</td>\n",
       "      <td>0.061528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-09-13</th>\n",
       "      <td>-0.086337</td>\n",
       "      <td>0.052149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-09-14</th>\n",
       "      <td>0.039840</td>\n",
       "      <td>-0.046887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-20</th>\n",
       "      <td>-0.032081</td>\n",
       "      <td>0.063270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-21</th>\n",
       "      <td>0.038743</td>\n",
       "      <td>0.031912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-22</th>\n",
       "      <td>-0.037849</td>\n",
       "      <td>0.027677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-23</th>\n",
       "      <td>0.044797</td>\n",
       "      <td>-0.016522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-24</th>\n",
       "      <td>-0.036049</td>\n",
       "      <td>0.014109</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>142 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Actual Log Diff Close  Predicted_Close\n",
       "date                                              \n",
       "2022-09-08               0.003289         0.000000\n",
       "2022-09-09               0.049426         0.000614\n",
       "2022-09-12              -0.001083         0.061528\n",
       "2022-09-13              -0.086337         0.052149\n",
       "2022-09-14               0.039840        -0.046887\n",
       "...                           ...              ...\n",
       "2023-03-20              -0.032081         0.063270\n",
       "2023-03-21               0.038743         0.031912\n",
       "2023-03-22              -0.037849         0.027677\n",
       "2023-03-23               0.044797        -0.016522\n",
       "2023-03-24              -0.036049         0.014109\n",
       "\n",
       "[142 rows x 2 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_df = pd.concat([log_diff_actual, log_diff_predicted], axis=1)\n",
    "prediction_df.rename(columns={'Log_Diff_Close': 'Actual Log Diff Close', 'Predicted Close': 'Predicted Log Diff Close'}, inplace=True)\n",
    "output_path = os.path.join(os.getcwd(), '../../data/output', \"prediction_\"+CLOSE+\".xlsx\")\n",
    "prediction_df.to_excel(output_path, index=False)\n",
    "\n",
    "prediction_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>rendement_predit</th>\n",
       "      <th>vol_empirique</th>\n",
       "      <th>prediction</th>\n",
       "      <th>alpha(expo_crypto)</th>\n",
       "      <th>rendement_observe</th>\n",
       "      <th>valeur_portefeuille</th>\n",
       "      <th>valeur_crypto</th>\n",
       "      <th>valeur_portefeuille_buy_and_hold</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-09-08</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>?</td>\n",
       "      <td>bear</td>\n",
       "      <td>100%</td>\n",
       "      <td>0.003289</td>\n",
       "      <td>10032.890636</td>\n",
       "      <td>1635.37</td>\n",
       "      <td>10032.890636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-09-09</td>\n",
       "      <td>0.000614</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>0.049426</td>\n",
       "      <td>10528.772404</td>\n",
       "      <td>1718.23</td>\n",
       "      <td>10528.772404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-09-12</td>\n",
       "      <td>0.061528</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>-0.001083</td>\n",
       "      <td>10517.368736</td>\n",
       "      <td>1716.37</td>\n",
       "      <td>10517.368736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-09-13</td>\n",
       "      <td>0.052149</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>-0.086337</td>\n",
       "      <td>9609.327009</td>\n",
       "      <td>1574.40</td>\n",
       "      <td>9609.327009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-09-14</td>\n",
       "      <td>-0.046887</td>\n",
       "      <td>?</td>\n",
       "      <td>bear</td>\n",
       "      <td>100%</td>\n",
       "      <td>0.039840</td>\n",
       "      <td>9226.493295</td>\n",
       "      <td>1638.39</td>\n",
       "      <td>9992.160723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>2023-03-20</td>\n",
       "      <td>0.063270</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>-0.032081</td>\n",
       "      <td>12119.446467</td>\n",
       "      <td>1732.78</td>\n",
       "      <td>9270.812194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>2023-03-21</td>\n",
       "      <td>0.031912</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>0.038743</td>\n",
       "      <td>12588.986663</td>\n",
       "      <td>1801.23</td>\n",
       "      <td>9629.988579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>2023-03-22</td>\n",
       "      <td>0.027677</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>-0.037849</td>\n",
       "      <td>12112.511228</td>\n",
       "      <td>1734.33</td>\n",
       "      <td>9265.507059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>2023-03-23</td>\n",
       "      <td>-0.016522</td>\n",
       "      <td>?</td>\n",
       "      <td>bear</td>\n",
       "      <td>100%</td>\n",
       "      <td>0.044797</td>\n",
       "      <td>11569.902132</td>\n",
       "      <td>1813.79</td>\n",
       "      <td>9680.577750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>2023-03-24</td>\n",
       "      <td>0.014109</td>\n",
       "      <td>?</td>\n",
       "      <td>bull</td>\n",
       "      <td>100%</td>\n",
       "      <td>-0.036049</td>\n",
       "      <td>11152.824107</td>\n",
       "      <td>1749.57</td>\n",
       "      <td>9331.607102</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>142 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           date  rendement_predit vol_empirique prediction alpha(expo_crypto)  \\\n",
       "0    2022-09-08          0.000000             ?       bear               100%   \n",
       "1    2022-09-09          0.000614             ?       bull               100%   \n",
       "2    2022-09-12          0.061528             ?       bull               100%   \n",
       "3    2022-09-13          0.052149             ?       bull               100%   \n",
       "4    2022-09-14         -0.046887             ?       bear               100%   \n",
       "..          ...               ...           ...        ...                ...   \n",
       "137  2023-03-20          0.063270             ?       bull               100%   \n",
       "138  2023-03-21          0.031912             ?       bull               100%   \n",
       "139  2023-03-22          0.027677             ?       bull               100%   \n",
       "140  2023-03-23         -0.016522             ?       bear               100%   \n",
       "141  2023-03-24          0.014109             ?       bull               100%   \n",
       "\n",
       "     rendement_observe  valeur_portefeuille  valeur_crypto  \\\n",
       "0             0.003289         10032.890636        1635.37   \n",
       "1             0.049426         10528.772404        1718.23   \n",
       "2            -0.001083         10517.368736        1716.37   \n",
       "3            -0.086337          9609.327009        1574.40   \n",
       "4             0.039840          9226.493295        1638.39   \n",
       "..                 ...                  ...            ...   \n",
       "137          -0.032081         12119.446467        1732.78   \n",
       "138           0.038743         12588.986663        1801.23   \n",
       "139          -0.037849         12112.511228        1734.33   \n",
       "140           0.044797         11569.902132        1813.79   \n",
       "141          -0.036049         11152.824107        1749.57   \n",
       "\n",
       "     valeur_portefeuille_buy_and_hold  \n",
       "0                        10032.890636  \n",
       "1                        10528.772404  \n",
       "2                        10517.368736  \n",
       "3                         9609.327009  \n",
       "4                         9992.160723  \n",
       "..                                ...  \n",
       "137                       9270.812194  \n",
       "138                       9629.988579  \n",
       "139                       9265.507059  \n",
       "140                       9680.577750  \n",
       "141                       9331.607102  \n",
       "\n",
       "[142 rows x 9 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "money = 10000\n",
    "money_buy_and_hold = 10000\n",
    "money_evolution = []\n",
    "money_evolution_buy_and_hold = []\n",
    "backtest_values = []\n",
    "for i in range(len(log_diff_predicted)):\n",
    "    money_buy_and_hold = money_buy_and_hold * (1 + log_diff_actual.iloc[i][0])\n",
    "    money_evolution_buy_and_hold.append(money_buy_and_hold)\n",
    "    if log_diff_predicted[i] >= 0:\n",
    "        money = money * (1 + log_diff_actual.iloc[i][0])\n",
    "    else:\n",
    "        money = money * (1 - log_diff_actual.iloc[i][0])\n",
    "    money_evolution.append(money)\n",
    "    cur_date = log_diff_actual.index[i].strftime(\"%Y-%m-%d\")\n",
    "    cur_rendement_predit = log_diff_predicted.iloc[i]\n",
    "    cur_vol_empirique = \"?\"\n",
    "    cur_prediction = \"bull\" if log_diff_predicted[i] > 0 else \"bear\"\n",
    "    cur_alpha = \"100%\"\n",
    "    cur_rendement_observe = log_diff_actual.iloc[i][0]\n",
    "    cur_valeur_portefeuille = money\n",
    "    cur_valeur_crypto = df.loc[cur_date, CLOSE]\n",
    "    backtest_values.append([cur_date, cur_rendement_predit, cur_vol_empirique, cur_prediction, cur_alpha, cur_rendement_observe, cur_valeur_portefeuille, cur_valeur_crypto, money_buy_and_hold])\n",
    "backtest = pd.DataFrame(backtest_values, columns=['date', 'rendement_predit', 'vol_empirique', 'prediction', 'alpha(expo_crypto)', 'rendement_observe', 'valeur_portefeuille', 'valeur_crypto', 'valeur_portefeuille_buy_and_hold'])\n",
    "output_path = os.path.join(os.getcwd(), '../../data/output', \"backtest_\"+CLOSE+\".xlsx\")\n",
    "backtest.to_excel(output_path, index=False)\n",
    "backtest\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Indicator Name</th>\n",
       "      <th>Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Mean Square Error</td>\n",
       "      <td>3.123756e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Root Mean Square Error</td>\n",
       "      <td>1.767415e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Mean Absolute Error</td>\n",
       "      <td>1.747732e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Root Mean Square Error (log)</td>\n",
       "      <td>5.500000e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Mean Absolute Error (log)</td>\n",
       "      <td>3.900000e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Mean Absolute Percentage Error</td>\n",
       "      <td>1.221011e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>U-Theil1 Statistic</td>\n",
       "      <td>3.356287e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>U-Theil2 Statistic</td>\n",
       "      <td>1.226000e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Diebold-Mariano Test</td>\n",
       "      <td>-7.937000e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Rendement Absolue</td>\n",
       "      <td>1.150000e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Rendement Absolue Buy and Hold</td>\n",
       "      <td>-6.700000e-02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Indicator Name         Value\n",
       "0                Mean Square Error  3.123756e+08\n",
       "1           Root Mean Square Error  1.767415e+04\n",
       "2              Mean Absolute Error  1.747732e+04\n",
       "3     Root Mean Square Error (log)  5.500000e-02\n",
       "4        Mean Absolute Error (log)  3.900000e-02\n",
       "5   Mean Absolute Percentage Error  1.221011e+03\n",
       "6               U-Theil1 Statistic  3.356287e+03\n",
       "7               U-Theil2 Statistic  1.226000e+01\n",
       "8             Diebold-Mariano Test -7.937000e+01\n",
       "9                Rendement Absolue  1.150000e-01\n",
       "10  Rendement Absolue Buy and Hold -6.700000e-02"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mse = mean_squared_error(actual_prices_aligned, predictions)\n",
    "\n",
    "# Calculate RMSE\n",
    "rmse = math.sqrt(mean_squared_error(actual_prices_aligned, predictions))\n",
    "\n",
    "# Calculate MAE\n",
    "mae = mean_absolute_error(actual_prices_aligned, predictions)\n",
    "\n",
    "# Calculate RMSE Log diff\n",
    "rmse_log = math.sqrt(mean_squared_error(log_diff_actual, log_diff_predicted))\n",
    "\n",
    "# Calculate MAE Log diff\n",
    "mae_log = mean_absolute_error(log_diff_actual, log_diff_predicted)\n",
    "\n",
    "# Convert predictions DataFrame column to a NumPy array and flatten\n",
    "predictions_array = predictions.iloc[:, 0].to_numpy().flatten()\n",
    "\n",
    "# Now calculate MAPE\n",
    "# Ensure actual_prices_aligned is also a NumPy array for consistency\n",
    "actual_prices_aligned_array = actual_prices_aligned.to_numpy()\n",
    "\n",
    "mape = np.mean(np.abs((actual_prices_aligned_array - predictions_array) / actual_prices_aligned_array)) * 100\n",
    "\n",
    "# Handle division by zero or NaN values\n",
    "mape = np.nan_to_num(mape)\n",
    "\n",
    "\n",
    "# Ensure actual_prices_aligned and predictions are correctly aligned and are numpy arrays\n",
    "actual_values = actual_prices_aligned.values\n",
    "predicted_values = predictions.values\n",
    "\n",
    "# U-Theil1\n",
    "u_theil1 = np.sqrt(np.sum(np.square(predicted_values - actual_values)) / np.sum(np.square(np.diff(actual_values))))\n",
    "\n",
    "# U-Theil2\n",
    "u_theil2 = (np.sqrt(np.mean(np.square(predicted_values - actual_values))) / np.sqrt(np.mean(np.square(actual_values))))\n",
    "\n",
    "\n",
    "# Calculate the Dielbod-Mariano\n",
    "# Error from your model\n",
    "model_errors = actual_prices_aligned_array - predictions_array\n",
    "\n",
    "# Error from a naive model (using previous value as prediction)\n",
    "naive_errors = actual_prices_aligned_array - np.roll(actual_prices_aligned_array, 1)\n",
    "naive_errors[0] = 0  # handle the first element which doesn't have a previous value\n",
    "# Difference in errors\n",
    "error_diff = model_errors - naive_errors\n",
    "\n",
    "# Mean difference in errors\n",
    "mean_diff = np.mean(error_diff)\n",
    "\n",
    "# Variance of the differences\n",
    "var_diff = np.var(error_diff)\n",
    "\n",
    "# Number of observations\n",
    "n = len(error_diff)\n",
    "\n",
    "# Calculate DM test statistic\n",
    "dm_stat = mean_diff / np.sqrt(var_diff / n)\n",
    "\n",
    "absolute_rendement = money_evolution[-1] / 10000 - 1\n",
    "\n",
    "absolute_rendement_buy_and_hold = money_evolution_buy_and_hold[-1] / 10000 - 1\n",
    "\n",
    "\n",
    "data = {\n",
    "    'Indicator Name': ['Mean Square Error','Root Mean Square Error', 'Mean Absolute Error','Root Mean Square Error (log)', 'Mean Absolute Error (log)', 'Mean Absolute Percentage Error',\n",
    "                       'U-Theil1 Statistic', 'U-Theil2 Statistic', 'Diebold-Mariano Test', 'Rendement Absolue', 'Rendement Absolue Buy and Hold'],\n",
    "    'Value': [mse, rmse, mae, rmse_log, mae_log, mape, u_theil1, u_theil2, dm_stat, absolute_rendement, absolute_rendement_buy_and_hold]\n",
    "}\n",
    "\n",
    "table = pd.DataFrame(data)\n",
    "table['Value'] = table['Value'].apply(lambda x: round(x, 3))\n",
    "\n",
    "\n",
    "output_path = os.path.join(os.getcwd(), '../../data/output', \"performance_\"+CLOSE+\".xlsx\")\n",
    "table.to_excel(output_path, index=False)\n",
    "table\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 8))\n",
    "plt.subplot(2, 1, 1)\n",
    "plt.plot(money_evolution_buy_and_hold, label='Buy and hold', color='blue')\n",
    "plt.plot(money_evolution, label='Applied strategy', color='red')\n",
    "plt.title(f'Current Holdings ({CURRENCY})')\n",
    "plt.ylabel('Money')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(2, 1, 2)\n",
    "plt.plot(log_diff_actual, label='Actual', color='blue')\n",
    "plt.plot(log_diff_predicted, label='Predicted', color='red')\n",
    "plt.title(f'Logarithmic Difference')\n",
    "plt.xlabel('Time')\n",
    "plt.ylabel('Log Difference')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}