RNN_LSTM_Implementation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "RNN_LSTM_Implementation.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPGQLxgwofsSQqkmU+rN8Ql",
      "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/ugururesin/Deep_Learning_Projects/blob/main/RNN_LSTM_Implementation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SLkU0g4UlyQR"
      },
      "source": [
        "### Content\n",
        "\n",
        "1. LSTM Implementation\n",
        "2. Visualize the results and convergence\n",
        "3. Multi layer LSTM implementation\n",
        "4. Variable sized inputs"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ftO5K4XkmCty"
      },
      "source": [
        "#### Toy Example"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3S8Wd7BomDFd"
      },
      "source": [
        "## IMPORTING LIBRARIRES\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "from keras.layers import LSTM\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1pzy_3hpmYxn"
      },
      "source": [
        "## DATA CREATION\n",
        "inputs = [[[i+j] for i in range(5)] for j in range(100)]\n",
        "target = [(i+5)  for i in range(100)]"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X1u8M-H3qCQS",
        "outputId": "43ff499c-b3f4-46cf-f541-7009d7423f35"
      },
      "source": [
        "inputs[0:5]"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[[[0], [1], [2], [3], [4]],\n",
              " [[1], [2], [3], [4], [5]],\n",
              " [[2], [3], [4], [5], [6]],\n",
              " [[3], [4], [5], [6], [7]],\n",
              " [[4], [5], [6], [7], [8]]]"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ql1cyAMKqJjI",
        "outputId": "f7592a32-6e20-415b-f57e-8ded7ed752e4"
      },
      "source": [
        "target[0:5]"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[5, 6, 7, 8, 9]"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0A8bG9etrUlz"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lqVNjKI9q68C"
      },
      "source": [
        "# DATA SPLITTING\n",
        "x_train, x_test, y_train, y_test = train_test_split(inputs , target, test_size=.2, random_state=4)"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dhLTF5Ybrc-q"
      },
      "source": [
        "# RNN MODEL\n",
        "model = Sequential()\n",
        "\n",
        "model.add(LSTM((1), batch_input_shape=(None,5,1), return_sequences=False))\n",
        "#1 is the OUTPUT-SIZE, batch_input_shape: #inputs, length_inputs, length_each vector\n",
        "\n",
        "model.compile(loss='mean_absolute_error', optimizer='adam', metrics=['accuracy'])"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0dZmbqjAsM67",
        "outputId": "23b7ab60-326f-4edc-98f8-ebd6cc4275d7"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "lstm (LSTM)                  (None, 1)                 12        \n",
            "=================================================================\n",
            "Total params: 12\n",
            "Trainable params: 12\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CNFxB1KRtHp7"
      },
      "source": [
        "history = model.fit(x_train, y_train, epochs=500, validation_data=(x_test, y_test), verbose=0)"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UMDAilH5tSQc"
      },
      "source": [
        "results = model.predict(x_test)"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "NMLO2sIKtmr8",
        "outputId": "9c81cdf6-5bc8-48dc-c104-f70a00ceeaea"
      },
      "source": [
        "plt.scatter(range(20), y_test, c='g')\n",
        "plt.scatter(range(20), results, c='r')\n",
        "plt.show()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "0s7Uceb_tw7m",
        "outputId": "f222de25-4fdf-4da1-b248-97498be39eef"
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.show()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uJfwT47zuW3u"
      },
      "source": [
        "As it can be seen on the plot, the loss is too high!  \n",
        "The main reason is that the data was not normalized!\n",
        "\n",
        "Thus, in order to increase the loss, **the data need to be normalized** first!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KetT2FHnut1T"
      },
      "source": [
        "## DATA NORMALIZATION\n",
        "normal_inputs = [[[(i+j)/100] for i in range(5)] for j in range(100)]\n",
        "normal_target = [(i+5)/100    for i in range(100)]"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NAHk2JcHutwe",
        "outputId": "2d238bf4-b288-4d59-a938-92b2c9a58ab5"
      },
      "source": [
        "normal_inputs[0:5]"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[[[0.0], [0.01], [0.02], [0.03], [0.04]],\n",
              " [[0.01], [0.02], [0.03], [0.04], [0.05]],\n",
              " [[0.02], [0.03], [0.04], [0.05], [0.06]],\n",
              " [[0.03], [0.04], [0.05], [0.06], [0.07]],\n",
              " [[0.04], [0.05], [0.06], [0.07], [0.08]]]"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B137s9_Xt8Fn",
        "outputId": "20210ba1-db8b-4784-8700-074f46a57265"
      },
      "source": [
        "# DATA SPLITTING\n",
        "x_train, x_test, y_train, y_test = train_test_split(normal_inputs , normal_target, test_size=.2, random_state=4)\n",
        "\n",
        "# RNN MODEL\n",
        "model = Sequential()\n",
        "\n",
        "model.add(LSTM((1), batch_input_shape=(None,5,1), return_sequences=False))\n",
        "#1 is the OUTPUT-SIZE, batch_input_shape: #inputs, length_inputs, length_each vector\n",
        "\n",
        "model.compile(loss='mean_absolute_error', optimizer='adam', metrics=['accuracy'])\n",
        "model.summary()\n",
        "\n",
        "history = model.fit(x_train, y_train, epochs=500, validation_data=(x_test, y_test), verbose=0)\n",
        "\n",
        "results = model.predict(x_test)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "lstm_3 (LSTM)                (None, 1)                 12        \n",
            "=================================================================\n",
            "Total params: 12\n",
            "Trainable params: 12\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "W1kKcas_v3Dq",
        "outputId": "ddf6aafe-0c3d-48e7-98bd-b6b786a8274c"
      },
      "source": [
        "plt.scatter(range(20), y_test, c='g')\n",
        "plt.scatter(range(20), results, c='r')\n",
        "plt.show()"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "23EcC8WLwEPm",
        "outputId": "f54f89cf-28c5-4ff2-b9d7-f15e557b6fd2"
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.show()"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-HexTdvBfDjz"
      },
      "source": [
        "### Variable Size Inputs  \n",
        "* Since the model is trained by inputs whose length is 5, we may padding zeros for the inputs whose length is less than 5.  \n",
        "\n",
        "* Another solution is to make the sequence **None** by making the **batch_input_size** (None, None, 1)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZxTT0w9sgQFr"
      },
      "source": [
        "## DATA NORMALIZATION\n",
        "normal_inputs = [[[(i+j)/100] for i in range(7)] for j in range(100)]\n",
        "normal_target = [(i+7)/100    for i in range(100)]\n",
        "\n",
        "# DATA SPLITTING\n",
        "x_train, x_test, y_train, y_test = train_test_split(normal_inputs , normal_target, test_size=.2, random_state=4)"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 480
        },
        "id": "RQmDKJMSfRiw",
        "outputId": "f4501c24-7057-4b9e-824a-b7c23f694ab1"
      },
      "source": [
        "# RNN MODEL\n",
        "model = Sequential()\n",
        "model.add(LSTM((1), batch_input_shape=(None,None,1), return_sequences=False))\n",
        "\n",
        "model.compile(loss='mean_absolute_error', optimizer='adam', metrics=['accuracy'])\n",
        "model.summary()\n",
        "\n",
        "history = model.fit(x_train, y_train, epochs=500, validation_data=(x_test, y_test), verbose=0)\n",
        "results = model.predict(x_test)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_8\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "lstm_8 (LSTM)                (None, 1)                 12        \n",
            "=================================================================\n",
            "Total params: 12\n",
            "Trainable params: 12\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "WARNING:tensorflow:6 out of the last 6 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7f3cf9f774d0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\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"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "URcCOQFeiswM",
        "outputId": "de484239-4813-4922-93a6-568ac790f173"
      },
      "source": [
        "plt.scatter(range(20), y_test, c='g')\n",
        "plt.scatter(range(20), results, c='r')\n",
        "plt.show()"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "id": "L14DuUQPir9o",
        "outputId": "f6e70989-47d3-4d28-dd9a-3e7899e7913d"
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.show()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}