regresionRNA14.10.2021.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "regresionYaiza.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1NjF9cLr8zCv5U-FHJM5K3WEaySpCHEKQ",
      "authorship_tag": "ABX9TyPz1W+4nzE734zFwTzlsKEH",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/angelminsua/python/blob/main/regresionRNA14.10.2021.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "CopIR19-XqJE",
        "outputId": "79013c97-5861-4871-986e-3d6c7c980c9a"
      },
      "source": [
        "\n",
        "from keras.models import Sequential\n",
        "from keras import layers\n",
        "import pathlib\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "#Multilayer Perceptron (MLP): https://www.pyimagesearch.com/2019/01/21/regression-with-keras/\n",
        "def create_mlp(dim, regress=False):\n",
        "\t# define our MLP network\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(layers.Dense(50, input_dim=dim, activation=\"relu\"))\n",
        "\tmodel.add(layers.Dense(25, activation=\"relu\"))\n",
        "\t# check to see if the regression node should be added\n",
        "\tif regress:\n",
        "\t  model.add(layers.Dense(1, activation=\"linear\"))\n",
        "\t\n",
        "\toptimizer = tf.keras.optimizers.RMSprop(0.001) #parametro que se puede cambiar para mejorar \n",
        "\tmodel.compile(loss='mse', optimizer=optimizer, metrics=['mae', 'mse'])\n",
        "\treturn model\n",
        "\n",
        "#alternativa al metodo anterior\n",
        "\"\"\"\n",
        "def build_model():\n",
        "  model = Sequential([layers.Dense(99, activation='relu', input_shape=[len(train_dataset.keys())]),layers.Dense(64, activation='relu'),layers.Dense(1)])\n",
        "\n",
        "  optimizer = tf.keras.optimizers.RMSprop(0.001)\n",
        "\n",
        "  model.compile(loss='mse',\n",
        "                optimizer=optimizer,\n",
        "                metrics=['mae', 'mse'])\n",
        "  return model\n",
        "\"\"\"\n",
        "\n",
        "print(tf.__version__)\n",
        "!pwd\n",
        "\n",
        "dataset = pd.read_excel('df_datos_regresion_googleColab.xlsx', sheet_name='Sheet1')\n",
        "print(\"dataset sin normalizar=\", dataset.head(3))\n",
        "\n",
        "# dividimos el set de datos en un set de entrenamiento y otro de pruebas. Usaremos el set de pruebas en la evaluacion final de nuestro modelo.\n",
        "\n",
        "\"\"\"\n",
        "train_dataset = dataset.sample(frac=0.8,random_state=0)  # train_data es de tipo DataFrame\n",
        "print(\"angel==================================================================\", len(train_dataset.keys()))\n",
        "test_data = dataset.drop(train_dataset.index) # test_data es de tipo DataFrame\n",
        "print(\"test_data=\", type(test_data))\n",
        "train_labels = train_dataset.pop('BusTime') #devuelve la columna BusTime en forma de objeto Series y la elimina del dataframe train_data\n",
        "#print(\"DF=======\", train_data.head(5))\n",
        "test_labels = test_data.pop('BusTime') #test_labels es de tipo Series. Hacemos lo mismo pero para el dataframe test_data (en la instrucción anterior fue para el dataframe train_data)\n",
        "print(\"test_labels=\", test_labels[:5])\n",
        "\"\"\"\n",
        "\n",
        "x_df = dataset.drop('BusTime', axis=1)\n",
        "y_df= dataset['BusTime']\n",
        "print(\"x_df=\", x_df)\n",
        "print(\"y_df=\", y_df)\n",
        "\n",
        "\n",
        "#normalizar dataset\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "scaler= MinMaxScaler()\n",
        "scaler = scaler.fit(x_df)\n",
        "x= scaler.transform(x_df)\n",
        "print(\"x_df=\", (x_df))\n",
        "y = y_df.to_numpy()\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "train_dataset, test_data, train_labels, test_labels = train_test_split(x, y, test_size=0.2, random_state = 0) #train_dataset es X_train, test_data es X_test, train_labels es y_train, test_labels es y_test \n",
        "print(\"train_dataset.shape[1]==========================================\", train_dataset.shape[1])\n",
        "\n",
        "model = create_mlp(train_dataset.shape[1],True) #build_model()\n",
        "model.summary()\n",
        "\n",
        "\n",
        "#Entrenar a la modelo\n",
        "#Entrene el modelo durante 1000 épocas y registre la precisión de entrenamiento y validación en el objeto history.\n",
        "\n",
        "# Display training progress by printing a single dot for each completed epoch\n",
        "class PrintDot(keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs):\n",
        "    if epoch % 100 == 0: print('')\n",
        "    print('.', end='')\n",
        "\n",
        "EPOCHS = 3000\n",
        "history = model.fit(train_dataset, train_labels, epochs=EPOCHS, validation_split = 0.2, verbose=0, callbacks=[PrintDot()])\n",
        "\n",
        "#Visualizar el progreso de entrenamiento del modelo usando las estadísticas almacenadas en el objeto history.\n",
        "hist = pd.DataFrame(history.history)\n",
        "hist['epoch'] = history.epoch\n",
        "hist.tail()\n",
        "\n",
        "\n",
        "hist = pd.DataFrame(history.history)\n",
        "hist['epoch'] = history.epoch\n",
        "#https://www.tensorflow.org/tutorials/keras/regression?hl=es-419\n",
        "plt.figure()\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Mean Abs Error [MPG]')\n",
        "plt.plot(hist['epoch'], hist['mae'],\n",
        "          label='Train Error')\n",
        "plt.plot(hist['epoch'], hist['val_mae'],\n",
        "          label = 'Val Error')\n",
        "plt.ylim([0,5])\n",
        "plt.legend()\n",
        "\n",
        "plt.figure()\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Mean Square Error [$MPG^2$]')\n",
        "plt.plot(hist['epoch'], hist['mse'],\n",
        "          label='Train Error')\n",
        "plt.plot(hist['epoch'], hist['val_mse'],\n",
        "          label = 'Val Error')\n",
        "plt.ylim([0,20])\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "\n",
        "#plot_history(history)\n",
        "\n",
        "loss, mae, mse = model.evaluate(test_data, test_labels, verbose=2)\n",
        "print(\"Testing set Mean Abs Error: {:5.2f} MPG\".format(mae))\n",
        "\n",
        "test_predictions = model.predict(test_data).flatten()\n",
        "#dibujamos las predicciones (cada punto debería dar a la misma altura en el eje x y en el y)\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.scatter(test_labels, test_predictions)\n",
        "plt.xlabel('True Values [MPG]')\n",
        "plt.ylabel('Predictions [MPG]')\n",
        "plt.axis('equal')\n",
        "plt.axis('square')\n",
        "plt.xlim([0,plt.xlim()[1]])\n",
        "plt.ylim([0,plt.ylim()[1]])\n",
        "_ = plt.plot([-100, 100], [-100, 100])\n",
        "plt.show()\n",
        "\n",
        "#Echemos un vistazo a la distribución de errores. Deberóa seguir una normal\n",
        "error = test_predictions - test_labels\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.hist(error, bins = 25)\n",
        "plt.xlabel(\"Prediction Error [MPG]\")\n",
        "_ = plt.ylabel(\"Count\")\n",
        "plt.show()\n",
        "\"\"\"\n",
        "from keras import backend as K\n",
        "def coeff_determination(y_true, y_pred):\n",
        "    from keras import backend as K\n",
        "    SS_res =  K.sum(K.square( y_true-y_pred ))\n",
        "    SS_tot = K.sum(K.square( y_true - K.mean(y_true) ) )\n",
        "    return ( 1 - SS_res/(SS_tot + K.epsilon()) )\n",
        "\n",
        "r_square = coeff_determination (test_labels, test_predictions)\n",
        "print(\"r_square =\", r_square )   \n",
        "\"\"\"\n",
        "\n",
        "from sklearn.metrics import r2_score\n",
        "r_square = r2_score(test_labels, test_predictions)\n",
        "print(\"r_square =\", r_square )"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.6.0\n",
            "/content\n",
            "dataset sin normalizar=    StreamPatronage  TrafficFlowRate  ...  PreviousPatronage  BusTime\n",
            "0                1       118.333333  ...                  6      504\n",
            "1               11       168.000000  ...                 11      743\n",
            "2               10       216.222222  ...                  9      632\n",
            "\n",
            "[3 rows x 7 columns]\n",
            "x_df=        StreamPatronage  TrafficFlowRate  ...  AccumulatedRainfall  PreviousPatronage\n",
            "0                    1       118.333333  ...                  0.0                  6\n",
            "1                   11       168.000000  ...                  0.0                 11\n",
            "2                   10       216.222222  ...                  0.0                  9\n",
            "3                   26       330.952381  ...                  0.0                 12\n",
            "4                   13       413.111111  ...                  0.0                 19\n",
            "...                ...              ...  ...                  ...                ...\n",
            "21586               20       467.222222  ...                  0.0                 24\n",
            "21587               20       390.000000  ...                  0.0                 22\n",
            "21588                1       408.000000  ...                  0.0                  9\n",
            "21589               11       383.555556  ...                  0.0                 19\n",
            "21590                1       265.833333  ...                  0.0                  8\n",
            "\n",
            "[21591 rows x 6 columns]\n",
            "y_df= 0        504\n",
            "1        743\n",
            "2        632\n",
            "3        785\n",
            "4        599\n",
            "        ... \n",
            "21586    622\n",
            "21587    694\n",
            "21588    480\n",
            "21589    659\n",
            "21590    476\n",
            "Name: BusTime, Length: 21591, dtype: int64\n",
            "x_df=        StreamPatronage  TrafficFlowRate  ...  AccumulatedRainfall  PreviousPatronage\n",
            "0                    1       118.333333  ...                  0.0                  6\n",
            "1                   11       168.000000  ...                  0.0                 11\n",
            "2                   10       216.222222  ...                  0.0                  9\n",
            "3                   26       330.952381  ...                  0.0                 12\n",
            "4                   13       413.111111  ...                  0.0                 19\n",
            "...                ...              ...  ...                  ...                ...\n",
            "21586               20       467.222222  ...                  0.0                 24\n",
            "21587               20       390.000000  ...                  0.0                 22\n",
            "21588                1       408.000000  ...                  0.0                  9\n",
            "21589               11       383.555556  ...                  0.0                 19\n",
            "21590                1       265.833333  ...                  0.0                  8\n",
            "\n",
            "[21591 rows x 6 columns]\n",
            "train_dataset.shape[1]========================================== 6\n",
            "Model: \"sequential_8\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_24 (Dense)             (None, 50)                350       \n",
            "_________________________________________________________________\n",
            "dense_25 (Dense)             (None, 25)                1275      \n",
            "_________________________________________________________________\n",
            "dense_26 (Dense)             (None, 1)                 26        \n",
            "=================================================================\n",
            "Total params: 1,651\n",
            "Trainable params: 1,651\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "....................................................................................................\n",
            "...................................................................................................."
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "135/135 - 0s - loss: 7143.5117 - mae: 66.8828 - mse: 7143.5117\n",
            "Testing set Mean Abs Error: 66.88 MPG\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "r_square = 0.6572530162585211\n"
          ]
        }
      ]
    }
  ]
}