examples/time_series/cases/8.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "467482e4",
   "metadata": {},
   "source": [
    "### Example 8\n",
    "\n",
    "Refer to [/examples/time_series/generate_data.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/generate_data.ipynb) for the artificial time-series dataset (used in this example) generation  \n",
    "\n",
    "|Case|Uncoupled All Sensors|Coupled All Sensors|Uncoupled Less Sensors|Coupled Less Sensors|\n",
    "|-|-|-|-|-|\n",
    "|Example Number|[2](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/2.ipynb)|8|[9](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/9.ipynb)|[11](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/11.ipynb)|\n",
    "|ID|0_1_0_3_128_1_0_1_0_0|1_1_0_3_128_1_0_1_0_0|0_1_0_2_128_1_0_1_0_0|1_1_0_2_128_1_0_1_0_0|\n",
    "|Total params|17,185|17,185|17,057|17,057|\n",
    "|Total time (min)|1.5|3.1|1.0|1.6|\n",
    "|Num of epochs|49|107|31|51|\n",
    "|Time per epoch (s)|1.79|1.76|1.92|1.85|\n",
    "|Train RMSE|0.53|0.57|1.60|1.42|\n",
    "|Test RMSE|0.58|0.81|1.68|1.58|\n",
    "\n",
    "- More difficult to do predictions when the sensors are coupled\n",
    "- As the number of sensors used in the input is decreased, the performance deteriorates as expected\n",
    "- The performance deteriorates less when the sensors are coupled as expected\n",
    "\n",
    "Previous example: [/examples/time_series/cases/7.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/7.ipynb)  \n",
    "Next example: [/examples/time_series/cases/9.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/9.ipynb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6a748aa8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.insert(0, '../') # To be able to reach 'functions.py'\n",
    "sys.path.insert(0, '../../../') # To be able to reach 'datasets' folder\n",
    "import functions as fn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f978cd32",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Do the settings in this cell and run all\n",
    "class Train_Sets:\n",
    "    vers = 1 # 0 or 1\n",
    "    within = 1 # 0 or 1\n",
    "    inter = 0 # 0 or 1\n",
    "    numOfSens = 3 # 1 or 2 or 3\n",
    "    batchSize = 128 # (2 ** 0) or ... or (2 ** 8)\n",
    "    useHalf = True # False or True\n",
    "    dense = False # False or True\n",
    "    shuffle = True # False or True\n",
    "    stateful = False # False or True\n",
    "    resetStates = False # False or True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2fb5b2fc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ID: 1_1_0_3_128_1_0_1_0_0\n"
     ]
    }
   ],
   "source": [
    "Train_Sets, Paths = fn.init(Train_Sets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e3350324",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape:      (9216, 32, 3)\n",
      "y_train shape:      (9216,)\n",
      "X_val shape:        (3072, 32, 3)\n",
      "y_val shape:        (3072,)\n",
      "X_test shape:       (3072, 32, 3)\n",
      "y_test shape:       (3072,)\n",
      "Common data type:   <class 'numpy.float32'>\n"
     ]
    }
   ],
   "source": [
    "X_train, X_val, X_test, y_train, y_val, y_test = fn.split_data(Train_Sets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6ad3409e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metal device set to: Apple M2\n",
      "\n",
      "systemMemory: 8.00 GB\n",
      "maxCacheSize: 2.67 GB\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 23:00:57.833463: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:306] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2022-12-11 23:00:57.833713: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:272] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/128\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 23:00:58.226080: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n",
      "2022-12-11 23:00:59.157929: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:00:59.329471: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:00:59.660797: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:00:59.947260: 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": [
      " 4/72 [>.............................] - ETA: 1s - loss: 33.5174 - root_mean_squared_error: 5.7894 - mean_absolute_error: 4.8922  "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 23:01:00.219075: 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": [
      "72/72 [==============================] - ETA: 0s - loss: 15.0350 - root_mean_squared_error: 3.8775 - mean_absolute_error: 3.0008"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 23:01:02.128180: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:01:02.193503: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:01:02.254419: 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": [
      "72/72 [==============================] - 4s 32ms/step - loss: 15.0350 - root_mean_squared_error: 3.8775 - mean_absolute_error: 3.0008 - val_loss: 5.2985 - val_root_mean_squared_error: 2.3018 - val_mean_absolute_error: 1.8034\n",
      "Epoch 2/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 4.6664 - root_mean_squared_error: 2.1602 - mean_absolute_error: 1.7027 - val_loss: 3.9881 - val_root_mean_squared_error: 1.9970 - val_mean_absolute_error: 1.5672\n",
      "Epoch 3/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 3.3219 - root_mean_squared_error: 1.8226 - mean_absolute_error: 1.4350 - val_loss: 2.9356 - val_root_mean_squared_error: 1.7134 - val_mean_absolute_error: 1.3159\n",
      "Epoch 4/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 2.7804 - root_mean_squared_error: 1.6675 - mean_absolute_error: 1.3085 - val_loss: 2.4963 - val_root_mean_squared_error: 1.5800 - val_mean_absolute_error: 1.2248\n",
      "Epoch 5/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 2.4938 - root_mean_squared_error: 1.5792 - mean_absolute_error: 1.2391 - val_loss: 2.5374 - val_root_mean_squared_error: 1.5929 - val_mean_absolute_error: 1.2399\n",
      "Epoch 6/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 2.2896 - root_mean_squared_error: 1.5131 - mean_absolute_error: 1.1879 - val_loss: 2.2054 - val_root_mean_squared_error: 1.4850 - val_mean_absolute_error: 1.1555\n",
      "Epoch 7/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 2.1520 - root_mean_squared_error: 1.4670 - mean_absolute_error: 1.1502 - val_loss: 2.2047 - val_root_mean_squared_error: 1.4848 - val_mean_absolute_error: 1.1631\n",
      "Epoch 8/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 2.0290 - root_mean_squared_error: 1.4244 - mean_absolute_error: 1.1166 - val_loss: 1.9895 - val_root_mean_squared_error: 1.4105 - val_mean_absolute_error: 1.0970\n",
      "Epoch 9/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.8887 - root_mean_squared_error: 1.3743 - mean_absolute_error: 1.0820 - val_loss: 1.9820 - val_root_mean_squared_error: 1.4078 - val_mean_absolute_error: 1.0986\n",
      "Epoch 10/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.8149 - root_mean_squared_error: 1.3472 - mean_absolute_error: 1.0608 - val_loss: 2.0091 - val_root_mean_squared_error: 1.4174 - val_mean_absolute_error: 1.1056\n",
      "Epoch 11/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.8034 - root_mean_squared_error: 1.3429 - mean_absolute_error: 1.0570 - val_loss: 1.7754 - val_root_mean_squared_error: 1.3324 - val_mean_absolute_error: 1.0409\n",
      "Epoch 12/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.6914 - root_mean_squared_error: 1.3006 - mean_absolute_error: 1.0229 - val_loss: 1.7664 - val_root_mean_squared_error: 1.3291 - val_mean_absolute_error: 1.0365\n",
      "Epoch 13/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.6267 - root_mean_squared_error: 1.2754 - mean_absolute_error: 1.0051 - val_loss: 1.6512 - val_root_mean_squared_error: 1.2850 - val_mean_absolute_error: 1.0087\n",
      "Epoch 14/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.5744 - root_mean_squared_error: 1.2547 - mean_absolute_error: 0.9879 - val_loss: 1.7163 - val_root_mean_squared_error: 1.3101 - val_mean_absolute_error: 1.0292\n",
      "Epoch 15/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.5457 - root_mean_squared_error: 1.2432 - mean_absolute_error: 0.9774 - val_loss: 1.7527 - val_root_mean_squared_error: 1.3239 - val_mean_absolute_error: 1.0568\n",
      "Epoch 16/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.4670 - root_mean_squared_error: 1.2112 - mean_absolute_error: 0.9526 - val_loss: 1.6891 - val_root_mean_squared_error: 1.2997 - val_mean_absolute_error: 1.0177\n",
      "Epoch 17/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.4466 - root_mean_squared_error: 1.2028 - mean_absolute_error: 0.9437 - val_loss: 1.5499 - val_root_mean_squared_error: 1.2449 - val_mean_absolute_error: 0.9689\n",
      "Epoch 18/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.4038 - root_mean_squared_error: 1.1848 - mean_absolute_error: 0.9300 - val_loss: 1.5343 - val_root_mean_squared_error: 1.2387 - val_mean_absolute_error: 0.9753\n",
      "Epoch 19/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.3407 - root_mean_squared_error: 1.1579 - mean_absolute_error: 0.9131 - val_loss: 1.5729 - val_root_mean_squared_error: 1.2542 - val_mean_absolute_error: 0.9702\n",
      "Epoch 20/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.3851 - root_mean_squared_error: 1.1769 - mean_absolute_error: 0.9303 - val_loss: 1.4995 - val_root_mean_squared_error: 1.2245 - val_mean_absolute_error: 0.9634\n",
      "Epoch 21/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.3492 - root_mean_squared_error: 1.1615 - mean_absolute_error: 0.9171 - val_loss: 1.4885 - val_root_mean_squared_error: 1.2200 - val_mean_absolute_error: 0.9547\n",
      "Epoch 22/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.2457 - root_mean_squared_error: 1.1161 - mean_absolute_error: 0.8790 - val_loss: 1.4236 - val_root_mean_squared_error: 1.1931 - val_mean_absolute_error: 0.9342\n",
      "Epoch 23/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.2172 - root_mean_squared_error: 1.1033 - mean_absolute_error: 0.8693 - val_loss: 1.3592 - val_root_mean_squared_error: 1.1658 - val_mean_absolute_error: 0.9061\n",
      "Epoch 24/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.2389 - root_mean_squared_error: 1.1130 - mean_absolute_error: 0.8751 - val_loss: 1.3336 - val_root_mean_squared_error: 1.1548 - val_mean_absolute_error: 0.9048\n",
      "Epoch 25/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.1876 - root_mean_squared_error: 1.0898 - mean_absolute_error: 0.8591 - val_loss: 1.3596 - val_root_mean_squared_error: 1.1660 - val_mean_absolute_error: 0.9122\n",
      "Epoch 26/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.1974 - root_mean_squared_error: 1.0943 - mean_absolute_error: 0.8591 - val_loss: 1.3368 - val_root_mean_squared_error: 1.1562 - val_mean_absolute_error: 0.9025\n",
      "Epoch 27/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.1340 - root_mean_squared_error: 1.0649 - mean_absolute_error: 0.8414 - val_loss: 1.3516 - val_root_mean_squared_error: 1.1626 - val_mean_absolute_error: 0.9050\n",
      "Epoch 28/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.1175 - root_mean_squared_error: 1.0571 - mean_absolute_error: 0.8353 - val_loss: 1.2370 - val_root_mean_squared_error: 1.1122 - val_mean_absolute_error: 0.8606\n",
      "Epoch 29/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.0964 - root_mean_squared_error: 1.0471 - mean_absolute_error: 0.8223 - val_loss: 1.3039 - val_root_mean_squared_error: 1.1419 - val_mean_absolute_error: 0.8809\n",
      "Epoch 30/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.0659 - root_mean_squared_error: 1.0324 - mean_absolute_error: 0.8186 - val_loss: 1.2709 - val_root_mean_squared_error: 1.1273 - val_mean_absolute_error: 0.8703\n",
      "Epoch 31/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.0871 - root_mean_squared_error: 1.0426 - mean_absolute_error: 0.8244 - val_loss: 1.5475 - val_root_mean_squared_error: 1.2440 - val_mean_absolute_error: 0.9956\n",
      "Epoch 32/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.0902 - root_mean_squared_error: 1.0441 - mean_absolute_error: 0.8267 - val_loss: 1.4726 - val_root_mean_squared_error: 1.2135 - val_mean_absolute_error: 0.9688\n",
      "Epoch 33/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.0260 - root_mean_squared_error: 1.0129 - mean_absolute_error: 0.8020 - val_loss: 1.2676 - val_root_mean_squared_error: 1.1259 - val_mean_absolute_error: 0.8849\n",
      "Epoch 34/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.0039 - root_mean_squared_error: 1.0019 - mean_absolute_error: 0.7893 - val_loss: 1.1981 - val_root_mean_squared_error: 1.0946 - val_mean_absolute_error: 0.8565\n",
      "Epoch 35/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9851 - root_mean_squared_error: 0.9925 - mean_absolute_error: 0.7863 - val_loss: 1.3158 - val_root_mean_squared_error: 1.1471 - val_mean_absolute_error: 0.8981\n",
      "Epoch 36/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9713 - root_mean_squared_error: 0.9856 - mean_absolute_error: 0.7777 - val_loss: 1.3008 - val_root_mean_squared_error: 1.1405 - val_mean_absolute_error: 0.9051\n",
      "Epoch 37/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9721 - root_mean_squared_error: 0.9859 - mean_absolute_error: 0.7771 - val_loss: 1.1183 - val_root_mean_squared_error: 1.0575 - val_mean_absolute_error: 0.8236\n",
      "Epoch 38/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9412 - root_mean_squared_error: 0.9702 - mean_absolute_error: 0.7677 - val_loss: 1.1620 - val_root_mean_squared_error: 1.0780 - val_mean_absolute_error: 0.8392\n",
      "Epoch 39/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9196 - root_mean_squared_error: 0.9589 - mean_absolute_error: 0.7571 - val_loss: 1.1990 - val_root_mean_squared_error: 1.0950 - val_mean_absolute_error: 0.8465\n",
      "Epoch 40/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.9586 - root_mean_squared_error: 0.9791 - mean_absolute_error: 0.7768 - val_loss: 1.0976 - val_root_mean_squared_error: 1.0477 - val_mean_absolute_error: 0.8120\n",
      "Epoch 41/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8944 - root_mean_squared_error: 0.9457 - mean_absolute_error: 0.7496 - val_loss: 1.1771 - val_root_mean_squared_error: 1.0849 - val_mean_absolute_error: 0.8328\n",
      "Epoch 42/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8744 - root_mean_squared_error: 0.9351 - mean_absolute_error: 0.7370 - val_loss: 1.0766 - val_root_mean_squared_error: 1.0376 - val_mean_absolute_error: 0.8039\n",
      "Epoch 43/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8593 - root_mean_squared_error: 0.9270 - mean_absolute_error: 0.7340 - val_loss: 1.0736 - val_root_mean_squared_error: 1.0361 - val_mean_absolute_error: 0.8028\n",
      "Epoch 44/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8386 - root_mean_squared_error: 0.9157 - mean_absolute_error: 0.7212 - val_loss: 1.0526 - val_root_mean_squared_error: 1.0259 - val_mean_absolute_error: 0.7972\n",
      "Epoch 45/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8100 - root_mean_squared_error: 0.9000 - mean_absolute_error: 0.7091 - val_loss: 1.0644 - val_root_mean_squared_error: 1.0317 - val_mean_absolute_error: 0.7982\n",
      "Epoch 46/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.8294 - root_mean_squared_error: 0.9107 - mean_absolute_error: 0.7198 - val_loss: 1.0281 - val_root_mean_squared_error: 1.0139 - val_mean_absolute_error: 0.7816\n",
      "Epoch 47/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7821 - root_mean_squared_error: 0.8844 - mean_absolute_error: 0.6975 - val_loss: 1.0223 - val_root_mean_squared_error: 1.0111 - val_mean_absolute_error: 0.7864\n",
      "Epoch 48/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7888 - root_mean_squared_error: 0.8881 - mean_absolute_error: 0.6998 - val_loss: 1.2061 - val_root_mean_squared_error: 1.0982 - val_mean_absolute_error: 0.8745\n",
      "Epoch 49/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7896 - root_mean_squared_error: 0.8886 - mean_absolute_error: 0.7005 - val_loss: 1.0703 - val_root_mean_squared_error: 1.0345 - val_mean_absolute_error: 0.8106\n",
      "Epoch 50/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7914 - root_mean_squared_error: 0.8896 - mean_absolute_error: 0.7016 - val_loss: 0.9993 - val_root_mean_squared_error: 0.9997 - val_mean_absolute_error: 0.7717\n",
      "Epoch 51/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7555 - root_mean_squared_error: 0.8692 - mean_absolute_error: 0.6874 - val_loss: 0.9449 - val_root_mean_squared_error: 0.9721 - val_mean_absolute_error: 0.7514\n",
      "Epoch 52/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7424 - root_mean_squared_error: 0.8616 - mean_absolute_error: 0.6818 - val_loss: 1.0050 - val_root_mean_squared_error: 1.0025 - val_mean_absolute_error: 0.7875\n",
      "Epoch 53/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7564 - root_mean_squared_error: 0.8697 - mean_absolute_error: 0.6879 - val_loss: 0.9199 - val_root_mean_squared_error: 0.9591 - val_mean_absolute_error: 0.7396\n",
      "Epoch 54/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7267 - root_mean_squared_error: 0.8525 - mean_absolute_error: 0.6732 - val_loss: 0.9366 - val_root_mean_squared_error: 0.9678 - val_mean_absolute_error: 0.7535\n",
      "Epoch 55/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.7077 - root_mean_squared_error: 0.8413 - mean_absolute_error: 0.6637 - val_loss: 0.9043 - val_root_mean_squared_error: 0.9509 - val_mean_absolute_error: 0.7334\n",
      "Epoch 56/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.7365 - root_mean_squared_error: 0.8582 - mean_absolute_error: 0.6791 - val_loss: 0.9439 - val_root_mean_squared_error: 0.9715 - val_mean_absolute_error: 0.7497\n",
      "Epoch 57/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.6955 - root_mean_squared_error: 0.8340 - mean_absolute_error: 0.6574 - val_loss: 0.9244 - val_root_mean_squared_error: 0.9615 - val_mean_absolute_error: 0.7473\n",
      "Epoch 58/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.6873 - root_mean_squared_error: 0.8290 - mean_absolute_error: 0.6544 - val_loss: 0.9430 - val_root_mean_squared_error: 0.9711 - val_mean_absolute_error: 0.7541\n",
      "Epoch 59/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6652 - root_mean_squared_error: 0.8156 - mean_absolute_error: 0.6416 - val_loss: 0.9669 - val_root_mean_squared_error: 0.9833 - val_mean_absolute_error: 0.7716\n",
      "Epoch 60/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.7022 - root_mean_squared_error: 0.8380 - mean_absolute_error: 0.6624 - val_loss: 0.8972 - val_root_mean_squared_error: 0.9472 - val_mean_absolute_error: 0.7390\n",
      "Epoch 61/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6767 - root_mean_squared_error: 0.8226 - mean_absolute_error: 0.6481 - val_loss: 0.9213 - val_root_mean_squared_error: 0.9599 - val_mean_absolute_error: 0.7496\n",
      "Epoch 62/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.6923 - root_mean_squared_error: 0.8321 - mean_absolute_error: 0.6556 - val_loss: 0.8854 - val_root_mean_squared_error: 0.9409 - val_mean_absolute_error: 0.7295\n",
      "Epoch 63/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.6671 - root_mean_squared_error: 0.8167 - mean_absolute_error: 0.6431 - val_loss: 0.8900 - val_root_mean_squared_error: 0.9434 - val_mean_absolute_error: 0.7399\n",
      "Epoch 64/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6496 - root_mean_squared_error: 0.8060 - mean_absolute_error: 0.6350 - val_loss: 0.8432 - val_root_mean_squared_error: 0.9183 - val_mean_absolute_error: 0.7098\n",
      "Epoch 65/128\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 0.6298 - root_mean_squared_error: 0.7936 - mean_absolute_error: 0.6269 - val_loss: 0.8346 - val_root_mean_squared_error: 0.9136 - val_mean_absolute_error: 0.7051\n",
      "Epoch 66/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6280 - root_mean_squared_error: 0.7925 - mean_absolute_error: 0.6265 - val_loss: 0.8721 - val_root_mean_squared_error: 0.9339 - val_mean_absolute_error: 0.7222\n",
      "Epoch 67/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.6476 - root_mean_squared_error: 0.8047 - mean_absolute_error: 0.6352 - val_loss: 0.8710 - val_root_mean_squared_error: 0.9333 - val_mean_absolute_error: 0.7212\n",
      "Epoch 68/128\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 0.6044 - root_mean_squared_error: 0.7774 - mean_absolute_error: 0.6117 - val_loss: 0.8281 - val_root_mean_squared_error: 0.9100 - val_mean_absolute_error: 0.7053\n",
      "Epoch 69/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.5956 - root_mean_squared_error: 0.7717 - mean_absolute_error: 0.6060 - val_loss: 0.8708 - val_root_mean_squared_error: 0.9332 - val_mean_absolute_error: 0.7338\n",
      "Epoch 70/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5949 - root_mean_squared_error: 0.7713 - mean_absolute_error: 0.6089 - val_loss: 0.7953 - val_root_mean_squared_error: 0.8918 - val_mean_absolute_error: 0.6917\n",
      "Epoch 71/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5968 - root_mean_squared_error: 0.7725 - mean_absolute_error: 0.6093 - val_loss: 0.8310 - val_root_mean_squared_error: 0.9116 - val_mean_absolute_error: 0.7090\n",
      "Epoch 72/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6024 - root_mean_squared_error: 0.7761 - mean_absolute_error: 0.6162 - val_loss: 0.8105 - val_root_mean_squared_error: 0.9003 - val_mean_absolute_error: 0.6994\n",
      "Epoch 73/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5875 - root_mean_squared_error: 0.7665 - mean_absolute_error: 0.6037 - val_loss: 0.8049 - val_root_mean_squared_error: 0.8972 - val_mean_absolute_error: 0.6929\n",
      "Epoch 74/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.6003 - root_mean_squared_error: 0.7748 - mean_absolute_error: 0.6083 - val_loss: 0.8402 - val_root_mean_squared_error: 0.9167 - val_mean_absolute_error: 0.7169\n",
      "Epoch 75/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5583 - root_mean_squared_error: 0.7472 - mean_absolute_error: 0.5875 - val_loss: 0.8059 - val_root_mean_squared_error: 0.8977 - val_mean_absolute_error: 0.6991\n",
      "Epoch 76/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5436 - root_mean_squared_error: 0.7373 - mean_absolute_error: 0.5791 - val_loss: 0.7974 - val_root_mean_squared_error: 0.8930 - val_mean_absolute_error: 0.6934\n",
      "Epoch 77/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5709 - root_mean_squared_error: 0.7556 - mean_absolute_error: 0.5923 - val_loss: 0.7867 - val_root_mean_squared_error: 0.8870 - val_mean_absolute_error: 0.6907\n",
      "Epoch 78/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5515 - root_mean_squared_error: 0.7426 - mean_absolute_error: 0.5834 - val_loss: 0.7507 - val_root_mean_squared_error: 0.8664 - val_mean_absolute_error: 0.6684\n",
      "Epoch 79/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5447 - root_mean_squared_error: 0.7380 - mean_absolute_error: 0.5833 - val_loss: 0.8114 - val_root_mean_squared_error: 0.9008 - val_mean_absolute_error: 0.6991\n",
      "Epoch 80/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5532 - root_mean_squared_error: 0.7438 - mean_absolute_error: 0.5866 - val_loss: 0.7647 - val_root_mean_squared_error: 0.8745 - val_mean_absolute_error: 0.6785\n",
      "Epoch 81/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5501 - root_mean_squared_error: 0.7417 - mean_absolute_error: 0.5832 - val_loss: 0.7746 - val_root_mean_squared_error: 0.8801 - val_mean_absolute_error: 0.6859\n",
      "Epoch 82/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5451 - root_mean_squared_error: 0.7383 - mean_absolute_error: 0.5816 - val_loss: 0.7589 - val_root_mean_squared_error: 0.8712 - val_mean_absolute_error: 0.6761\n",
      "Epoch 83/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5250 - root_mean_squared_error: 0.7246 - mean_absolute_error: 0.5729 - val_loss: 0.7492 - val_root_mean_squared_error: 0.8656 - val_mean_absolute_error: 0.6728\n",
      "Epoch 84/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5165 - root_mean_squared_error: 0.7187 - mean_absolute_error: 0.5670 - val_loss: 0.7473 - val_root_mean_squared_error: 0.8645 - val_mean_absolute_error: 0.6719\n",
      "Epoch 85/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5260 - root_mean_squared_error: 0.7253 - mean_absolute_error: 0.5708 - val_loss: 0.7597 - val_root_mean_squared_error: 0.8716 - val_mean_absolute_error: 0.6795\n",
      "Epoch 86/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5185 - root_mean_squared_error: 0.7200 - mean_absolute_error: 0.5640 - val_loss: 0.7257 - val_root_mean_squared_error: 0.8519 - val_mean_absolute_error: 0.6587\n",
      "Epoch 87/128\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 0.5321 - root_mean_squared_error: 0.7295 - mean_absolute_error: 0.5744 - val_loss: 0.7234 - val_root_mean_squared_error: 0.8505 - val_mean_absolute_error: 0.6559\n",
      "Epoch 88/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5019 - root_mean_squared_error: 0.7085 - mean_absolute_error: 0.5555 - val_loss: 0.7618 - val_root_mean_squared_error: 0.8728 - val_mean_absolute_error: 0.6830\n",
      "Epoch 89/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5040 - root_mean_squared_error: 0.7099 - mean_absolute_error: 0.5575 - val_loss: 0.7370 - val_root_mean_squared_error: 0.8585 - val_mean_absolute_error: 0.6655\n",
      "Epoch 90/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.5003 - root_mean_squared_error: 0.7073 - mean_absolute_error: 0.5568 - val_loss: 0.7173 - val_root_mean_squared_error: 0.8469 - val_mean_absolute_error: 0.6578\n",
      "Epoch 91/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4985 - root_mean_squared_error: 0.7061 - mean_absolute_error: 0.5560 - val_loss: 0.7666 - val_root_mean_squared_error: 0.8756 - val_mean_absolute_error: 0.6794\n",
      "Epoch 92/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4973 - root_mean_squared_error: 0.7052 - mean_absolute_error: 0.5523 - val_loss: 0.7235 - val_root_mean_squared_error: 0.8506 - val_mean_absolute_error: 0.6595\n",
      "Epoch 93/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4984 - root_mean_squared_error: 0.7059 - mean_absolute_error: 0.5516 - val_loss: 0.7122 - val_root_mean_squared_error: 0.8439 - val_mean_absolute_error: 0.6541\n",
      "Epoch 94/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4736 - root_mean_squared_error: 0.6882 - mean_absolute_error: 0.5416 - val_loss: 0.7447 - val_root_mean_squared_error: 0.8629 - val_mean_absolute_error: 0.6660\n",
      "Epoch 95/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4760 - root_mean_squared_error: 0.6899 - mean_absolute_error: 0.5435 - val_loss: 0.7192 - val_root_mean_squared_error: 0.8480 - val_mean_absolute_error: 0.6565\n",
      "Epoch 96/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4973 - root_mean_squared_error: 0.7052 - mean_absolute_error: 0.5532 - val_loss: 0.8809 - val_root_mean_squared_error: 0.9386 - val_mean_absolute_error: 0.7427\n",
      "Epoch 97/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4723 - root_mean_squared_error: 0.6872 - mean_absolute_error: 0.5424 - val_loss: 0.7734 - val_root_mean_squared_error: 0.8794 - val_mean_absolute_error: 0.6823\n",
      "Epoch 98/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4651 - root_mean_squared_error: 0.6820 - mean_absolute_error: 0.5374 - val_loss: 0.7341 - val_root_mean_squared_error: 0.8568 - val_mean_absolute_error: 0.6691\n",
      "Epoch 99/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4750 - root_mean_squared_error: 0.6892 - mean_absolute_error: 0.5390 - val_loss: 0.6636 - val_root_mean_squared_error: 0.8146 - val_mean_absolute_error: 0.6287\n",
      "Epoch 100/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4739 - root_mean_squared_error: 0.6884 - mean_absolute_error: 0.5399 - val_loss: 0.7707 - val_root_mean_squared_error: 0.8779 - val_mean_absolute_error: 0.6860\n",
      "Epoch 101/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4437 - root_mean_squared_error: 0.6661 - mean_absolute_error: 0.5220 - val_loss: 0.6963 - val_root_mean_squared_error: 0.8344 - val_mean_absolute_error: 0.6471\n",
      "Epoch 102/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4476 - root_mean_squared_error: 0.6690 - mean_absolute_error: 0.5254 - val_loss: 0.7005 - val_root_mean_squared_error: 0.8370 - val_mean_absolute_error: 0.6527\n",
      "Epoch 103/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4573 - root_mean_squared_error: 0.6762 - mean_absolute_error: 0.5308 - val_loss: 0.7008 - val_root_mean_squared_error: 0.8371 - val_mean_absolute_error: 0.6510\n",
      "Epoch 104/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4444 - root_mean_squared_error: 0.6666 - mean_absolute_error: 0.5216 - val_loss: 0.6959 - val_root_mean_squared_error: 0.8342 - val_mean_absolute_error: 0.6454\n",
      "Epoch 105/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4338 - root_mean_squared_error: 0.6586 - mean_absolute_error: 0.5148 - val_loss: 0.6748 - val_root_mean_squared_error: 0.8215 - val_mean_absolute_error: 0.6391\n",
      "Epoch 106/128\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4288 - root_mean_squared_error: 0.6549 - mean_absolute_error: 0.5143 - val_loss: 0.7198 - val_root_mean_squared_error: 0.8484 - val_mean_absolute_error: 0.6618\n",
      "Epoch 107/128\n",
      "70/72 [============================>.] - ETA: 0s - loss: 0.4541 - root_mean_squared_error: 0.6738 - mean_absolute_error: 0.5310Restoring model weights from the end of the best epoch: 99.\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 0.4556 - root_mean_squared_error: 0.6750 - mean_absolute_error: 0.5312 - val_loss: 0.6947 - val_root_mean_squared_error: 0.8335 - val_mean_absolute_error: 0.6487\n",
      "Epoch 107: early stopping\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as lstm_cell_layer_call_fn, lstm_cell_layer_call_and_return_conditional_losses, lstm_cell_1_layer_call_fn, lstm_cell_1_layer_call_and_return_conditional_losses while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /Users/serhatsoyer/Repos/py4ML/datasets/time_series/cases/1_1_0_3_128_1_0_1_0_0/model/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /Users/serhatsoyer/Repos/py4ML/datasets/time_series/cases/1_1_0_3_128_1_0_1_0_0/model/assets\n"
     ]
    }
   ],
   "source": [
    "model, df, train_time = fn.get_model(Train_Sets, X_train, X_val, y_train, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1c649667",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"1_1_0_3_128_1_0_1_0_0\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm1 (LSTM)                (128, 32, 32)             4608      \n",
      "                                                                 \n",
      " lstm2 (LSTM)                (128, 32)                 8320      \n",
      "                                                                 \n",
      " dense3 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " dense4 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " dense5 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " drop0.0625 (Dropout)        (128, 32)                 0         \n",
      "                                                                 \n",
      " dense6 (Dense)              (128, 32)                 1056      \n",
      "                                                                 \n",
      " out (Dense)                 (128, 1)                  33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 17,185\n",
      "Trainable params: 17,185\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6a3689ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total training time:          3.1 min\n",
      "Num of epochs:                107\n",
      "Best val RMSE epoch:          99\n",
      "Training time per epoch:      1.76 sec\n"
     ]
    }
   ],
   "source": [
    "fn.train_info(df, train_time)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "9939f067",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fn.plot_learn_curve(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fc576e3a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-11 23:04:13.690805: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:04:13.747967: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n",
      "2022-12-11 23:04:13.816329: 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": [
      "y_train_pred shape: (9216,)\n",
      "y_val_pred shape:   (3072,)\n",
      "y_test_pred shape:  (3072,)\n",
      "Common data type:   <class 'numpy.float32'>\n"
     ]
    }
   ],
   "source": [
    "y_train_pred, y_val_pred, y_test_pred = fn.get_predictions(Train_Sets, model, X_train, X_val, X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8c28b462",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mse_train:     0.33\n",
      "mse_val:       0.66\n",
      "mse_test:      0.66\n",
      "rmse_train:    0.57\n",
      "rmse_val:      0.81\n",
      "rmse_test:     0.81\n",
      "mae_train:     0.45\n",
      "mae_val:       0.63\n",
      "mae_test:      0.63\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fn.evaluate_model(y_train, y_val, y_test, y_train_pred, y_val_pred, y_test_pred)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1a17012b",
   "metadata": {},
   "source": [
    "Previous example: [/examples/time_series/cases/7.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/7.ipynb)  \n",
    "Next example: [/examples/time_series/cases/9.ipynb](https://github.com/serhatsoyer/py4ML/blob/main/examples/time_series/cases/9.ipynb)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.6 ('study')",
   "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": "3.10.8 (main, Nov 24 2022, 08:08:27) [Clang 14.0.6 ]"
  },
  "vscode": {
   "interpreter": {
    "hash": "29d06b973d1ddb34db3279b24f9b5152402e688db937648b736e855ec4de60c0"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}