main.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.7.3"
    },
    "colab": {
      "name": "llll.ipynb",
      "provenance": [],
      "toc_visible": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZYibEQQ8SDSF",
        "colab_type": "text"
      },
      "source": [
        "# Libraries in use"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QpmwbSSnSDSM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#%pylab inline\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "plt.style.use('seaborn')\n",
        "#%matplotlib inline\n",
        "from datetime import datetime\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "import keras\n",
        "from keras.layers import Dense, Dropout\n",
        "from keras.models import Sequential\n",
        "from sklearn import metrics"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fUVEd7noSDSb",
        "colab_type": "text"
      },
      "source": [
        "# Data Preview"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y6Rqbs1_SDSe",
        "colab_type": "code",
        "outputId": "5266d0f5-c71a-4e86-f5d3-8eb41bc96098",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 528
        }
      },
      "source": [
        "dataset = pd.read_csv('IRO1NIKI0001_165288034.csv')\n",
        "dataset.drop(['Date','Time'], axis=1, inplace=True)\n",
        "plt.plot(dataset[\"Close\"])\n",
        "dataset.head(5)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Open</th>\n",
              "      <th>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1342</td>\n",
              "      <td>1412</td>\n",
              "      <td>1335</td>\n",
              "      <td>1335</td>\n",
              "      <td>8688194</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1343</td>\n",
              "      <td>1344</td>\n",
              "      <td>1322</td>\n",
              "      <td>1328</td>\n",
              "      <td>3869651</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1355</td>\n",
              "      <td>1377</td>\n",
              "      <td>1340</td>\n",
              "      <td>1366</td>\n",
              "      <td>1405458</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1349</td>\n",
              "      <td>1370</td>\n",
              "      <td>1347</td>\n",
              "      <td>1360</td>\n",
              "      <td>806186</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1350</td>\n",
              "      <td>1365</td>\n",
              "      <td>1349</td>\n",
              "      <td>1358</td>\n",
              "      <td>575378</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Open  High   Low  Close   Volume\n",
              "0  1342  1412  1335   1335  8688194\n",
              "1  1343  1344  1322   1328  3869651\n",
              "2  1355  1377  1340   1366  1405458\n",
              "3  1349  1370  1347   1360   806186\n",
              "4  1350  1365  1349   1358   575378"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAFKCAYAAADScRzUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dWizqp1cSDSn",
        "colab_type": "text"
      },
      "source": [
        "# Data Preprocess And Make It Smoother"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bSEvuW1USDSp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class DataPreprocessor:\n",
        "    # def __init__(self):\n",
        "    #     pass\n",
        "\n",
        "    def CustomSmoother(self, x, alpha = 0.2):  \n",
        "        s0 = x[0]\n",
        "        smoothed_statistic = [s0]\n",
        "        n = x.shape[0]\n",
        "        for i in range(1, n):\n",
        "            s1 = alpha * x[i] + (1 - alpha) * s0\n",
        "            smoothed_statistic.append(s1)\n",
        "            s0 = s1\n",
        "        smoothed_statistic = np.array(smoothed_statistic)\n",
        "        return pd.DataFrame(smoothed_statistic)\n",
        "\n",
        "    def PandaSmoother(self, x):\n",
        "        return pd.DataFrame(x).ewm(span=20).mean()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jqyo9dbKSDSw",
        "colab_type": "code",
        "outputId": "aa169020-a412-4dc2-bb66-cd711c7b11f5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 366
        }
      },
      "source": [
        "smoother = DataPreprocessor()\n",
        "# ohclv_data = np.c_[dataset['Open'],\n",
        "#                    dataset['High'],\n",
        "#                    dataset['Low'],\n",
        "#                    dataset['Close'],\n",
        "#                    dataset['Volume']]\n",
        "smoothened_ohclv_data = smoother.CustomSmoother(dataset.values)\n",
        "smoothened_ohclv_data1 = smoother.PandaSmoother(dataset.values)\n",
        "\n",
        "#pd.set_option('display.max_rows', 500)\n",
        "plt.plot(dataset[\"Close\"])\n",
        "plt.plot(smoothened_ohclv_data[3])\n",
        "plt.plot(smoothened_ohclv_data1[3])\n",
        "\n",
        "#smoothened_ohclv_data.head(100)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fc42617c5f8>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAFKCAYAAADScRzUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neT0AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h2v6hWerSDS7",
        "colab_type": "text"
      },
      "source": [
        "# Get Prepared Data With Technical Indicatos As Features\n",
        "### Indicators are included:\n",
        "### Exponential Weighted Moving Average (EMA), Relative Strength Index (RSI), Stochastic Oscillator, Williams, Moving Average Convergence Divergence (MACD),Price Rate Of Change, On Balance Volume"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TKJuCntdSDS_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def ema(x, p):\n",
        "    #x = x.squeeze()\n",
        "    prev_ema = x[:p].mean()\n",
        "    ema = [prev_ema]\n",
        "    m = len(x)\n",
        "    multiplier = 2 / float(p + 1)\n",
        "    for i in range(p, m):\n",
        "        cur_ema = (x[i] - prev_ema) * multiplier + prev_ema\n",
        "        prev_ema = cur_ema\n",
        "        ema.append(cur_ema)\n",
        "    return np.array(ema)\n",
        "\n",
        "\n",
        "def getRSI(x):\n",
        "    #x = x.squeeze()\n",
        "    n = len(x)\n",
        "    x0 = x[:n - 1]\n",
        "    x1 = x[1:]\n",
        "    change = x1 - x0\n",
        "    avgGain = []\n",
        "    avgLoss = []\n",
        "    loss = 0\n",
        "    gain = 0\n",
        "    for i in range(14):\n",
        "        if change[i] > 0:\n",
        "            gain += change[i]\n",
        "        elif change[i] < 0:\n",
        "            loss += abs(change[i])\n",
        "    averageGain = gain / 14.0\n",
        "    averageLoss = loss / 14.0\n",
        "    avgGain.append(averageGain)\n",
        "    avgLoss.append(averageLoss)\n",
        "    for i in range(14, n - 1):\n",
        "        if change[i] >= 0:\n",
        "            avgGain.append((avgGain[-1] * 13 + change[i]) / 14.0)\n",
        "            avgLoss.append((avgLoss[-1] * 13) / 14.0)\n",
        "        else:\n",
        "            avgGain.append((avgGain[-1] * 13) / 14.0)\n",
        "            avgLoss.append((avgLoss[-1] * 13 + abs(change[i])) / 14.0)\n",
        "    avgGain = np.array(avgGain)\n",
        "    avgLoss = np.array(avgLoss)\n",
        "    RS = avgGain / avgLoss\n",
        "    RSI = 100 - (100 / (1 + RS))\n",
        "    return np.c_[RSI, x1[13:]]\n",
        "\n",
        "\n",
        "def getStochasticOscillator(x):\n",
        "    high = x[:, 1].squeeze()\n",
        "    low = x[:, 2].squeeze()\n",
        "    close = x[:, 3].squeeze()\n",
        "    n = len(high)\n",
        "    highestHigh = []\n",
        "    lowestLow = []\n",
        "    for i in range(n - 13):\n",
        "        highestHigh.append(high[i:i + 14].max())\n",
        "        lowestLow.append(low[i:i + 14].min())\n",
        "    highestHigh = np.array(highestHigh)\n",
        "    lowestLow = np.array(lowestLow)\n",
        "    k = 100 * ((close[13:] - lowestLow) / (highestHigh - lowestLow))\n",
        "\n",
        "    return np.c_[k, close[13:]]\n",
        "\n",
        "\n",
        "def getWilliams(x):\n",
        "    high = x[:, 1].squeeze()\n",
        "    low = x[:, 2].squeeze()\n",
        "    close = x[:, 3].squeeze()\n",
        "    n = len(high)\n",
        "    highestHigh = []\n",
        "    lowestLow = []\n",
        "    for i in range(n - 13):\n",
        "        highestHigh.append(high[i:i + 14].max())\n",
        "        lowestLow.append(low[i:i + 14].min())\n",
        "    highestHigh = np.array(highestHigh)\n",
        "    lowestLow = np.array(lowestLow)\n",
        "    w = -100 * ((highestHigh - close[13:]) / (highestHigh - lowestLow))\n",
        "    return np.c_[w, close[13:]]\n",
        "\n",
        "\n",
        "def getMACD(close):\n",
        "    ma1 = ema(close.squeeze(), 12)\n",
        "    ma2 = ema(close.squeeze(), 26)\n",
        "    macd = ma1[14:] - ma2\n",
        "    return np.c_[macd, close[len(close) - len(macd):]]\n",
        "\n",
        "\n",
        "def getPriceRateOfChange(close, n_days):\n",
        "    close = close.squeeze()\n",
        "    n = len(close)\n",
        "    x0 = close[:n - n_days]\n",
        "    x1 = close[n_days:]\n",
        "    PriceRateOfChange = (x1 - x0) / x0\n",
        "    return np.c_[PriceRateOfChange, x1]\n",
        "\n",
        "\n",
        "def getOnBalanceVolume(X):\n",
        "    close = X[:, 3].squeeze()\n",
        "    volume = X[:, 4].squeeze()[1:]\n",
        "    n = len(close)\n",
        "    x0 = close[:n - 1]\n",
        "    x1 = close[1:]\n",
        "    change = x1 - x0\n",
        "    OBV = []\n",
        "    prev_OBV = 0\n",
        "\n",
        "    for i in range(n - 1):\n",
        "        if change[i] > 0:\n",
        "            current_OBV = prev_OBV + volume[i]\n",
        "        elif change[i] < 0:\n",
        "            current_OBV = prev_OBV - volume[i]\n",
        "        else:\n",
        "            current_OBV = prev_OBV\n",
        "        OBV.append(current_OBV)\n",
        "        prev_OBV = current_OBV\n",
        "    OBV = np.array(OBV)\n",
        "    return np.c_[OBV, x1]\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y54N4eIaSDTK",
        "colab_type": "code",
        "outputId": "400b36d0-e7f4-4d49-a001-aff2dc6422a8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 481
        }
      },
      "source": [
        "days = 90\n",
        "ohclv_data = np.array(smoothened_ohclv_data)\n",
        "\n",
        "RSI = getRSI(ohclv_data[:, 3])\n",
        "# print(ohclv_data[:, 3])\n",
        "# print(\"kkkk\")\n",
        "# print(RSI)\n",
        "StochasticOscillator = getStochasticOscillator(ohclv_data)\n",
        "Williams = getWilliams(ohclv_data)\n",
        "\n",
        "MACD = getMACD(ohclv_data[:, 3])\n",
        "PROC = getPriceRateOfChange(ohclv_data[:, 3], days)\n",
        "OBV = getOnBalanceVolume(ohclv_data)\n",
        "\n",
        "min_len = min(len(RSI),\n",
        "              len(StochasticOscillator),\n",
        "              len(Williams),\n",
        "              len(MACD),\n",
        "              len(PROC),\n",
        "              len(OBV))\n",
        "\n",
        "RSI = RSI[len(RSI) - min_len:]\n",
        "StochasticOscillator = StochasticOscillator[len(StochasticOscillator) - min_len:]\n",
        "Williams = Williams[len(Williams) - min_len:]\n",
        "MACD = MACD[len(MACD) - min_len:]\n",
        "PROC = PROC[len(PROC) - min_len:]\n",
        "OBV = OBV[len(OBV) - min_len:]\n",
        "\n",
        "feature_matrix = np.c_[RSI[:, 0],\n",
        "                       StochasticOscillator[:, 0],\n",
        "                       Williams[:, 0],\n",
        "                       MACD[:, 0],\n",
        "                       PROC[:, 0],\n",
        "                       OBV[:, 0]]\n",
        "\n",
        "print(\"Size of feature matrix: \", feature_matrix.shape)\n",
        "number_of_samples = feature_matrix.shape[0]\n",
        "print(\"Number of samples: %d\" % number_of_samples)\n",
        "close = np.array(dataset['Close'])\n",
        "close = close[len(close) - number_of_samples:] #90 taye aval hazf shode\n",
        "y0 = close[:number_of_samples - days]\n",
        "y1 = close[days:]\n",
        "print(y0[:10])\n",
        "print(y1[:10])\n",
        "\n",
        "feature_matrix = feature_matrix[:number_of_samples - days]\n",
        "y = np.sign(y1 - y0)\n",
        "y[y == 0] = 1\n",
        "plt.hist(y)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Size of feature matrix:  (808, 6)\n",
            "Number of samples: 808\n",
            "[1895 1821 1816 1759 1771 1825 1741 1775 1807 1848]\n",
            "[1710 1708 1700 1738 1710 1714 1775 1792 1730 1745]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([332.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0., 386.]),\n",
              " array([-1. , -0.8, -0.6, -0.4, -0.2,  0. ,  0.2,  0.4,  0.6,  0.8,  1. ]),\n",
              " <a list of 10 Patch objects>)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeQAAAFKCAYAAADMuCxnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FXzIZy4fSDTQ",
        "colab_type": "text"
      },
      "source": [
        "# Data random train and test subsets and normalization"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LcqiFQMVSDTU",
        "colab_type": "code",
        "outputId": "e75a7bc6-84e4-4739-81b4-117857273f73",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        }
      },
      "source": [
        "X, X_test, y, y_test = train_test_split(feature_matrix, y) #random joda mikone bara test\n",
        "print(X.shape)\n",
        "# X = np.array(X).reshape(-1,6)\n",
        "# print(X.shape)\n",
        "y = np.array(y).reshape(-1,1)\n",
        "X_test = np.array(X_test).reshape(-1,6)\n",
        "y_test = np.array(y_test).reshape(-1,1)\n",
        "\n",
        "x_scaler = MinMaxScaler()\n",
        "y_scaler = MinMaxScaler() #adad ro beine sefr o yek mikone\n",
        "y = y_scaler.fit_transform(y)\n",
        "X = x_scaler.fit_transform(X)\n",
        "# print(X[0])\n",
        "\n",
        "# sigmoid = 1 / (1 + np.exp(-X[0]))\n",
        "# print(sigmoid)\n",
        "\n",
        "X_test = x_scaler.fit_transform(X_test)\n",
        "y_test = y_scaler.fit_transform(y_test)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(538, 6)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hmjPeqF5SDTd",
        "colab_type": "text"
      },
      "source": [
        "# 1) Define Network And Train It With Keras library"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3ziPoPbMSDTf",
        "colab_type": "code",
        "outputId": "6ce6a1d5-da3e-4830-8d41-1b138b1e23ce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 114
        }
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Dense(6, activation='relu'))\n",
        "model.add(Dense(10, activation='relu'))\n",
        "model.add(Dense(10, activation='relu'))\n",
        "model.add(Dense(10, activation='relu'))\n",
        "model.add(Dense(1, activation='linear'))\n",
        "\n",
        "model.compile(loss='mse', optimizer=keras.optimizers.SGD(lr=0.01), metrics=['acc'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ooJRRsg0SDTt",
        "colab_type": "code",
        "outputId": "331de611-f895-458e-a492-9f01366037f3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 438
        }
      },
      "source": [
        "history = model.fit(X,y,epochs=500,batch_size=32,validation_split=0.2,verbose=0)\n",
        "score = model.evaluate(X_test, y_test, verbose=0)\n",
        "Y_PREDICT = model.predict(X_test, batch_size=32)\n",
        "print(score)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1033: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1020: The name tf.assign is deprecated. Please use tf.compat.v1.assign instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3005: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:190: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:197: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:207: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:216: The name tf.is_variable_initialized is deprecated. Please use tf.compat.v1.is_variable_initialized instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:223: The name tf.variables_initializer is deprecated. Please use tf.compat.v1.variables_initializer instead.\n",
            "\n",
            "[0.12484388020303515, 0.8388888915379842]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IoVXCQo3SDT0",
        "colab_type": "text"
      },
      "source": [
        "# Calculate All Metrics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0GQWgt5tSDT5",
        "colab_type": "code",
        "outputId": "0b3f28f9-1166-4070-d92f-6b4dec128323",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 94
        }
      },
      "source": [
        "Y_PREDICT = [1 if x > 0.5 else 0 for x in Y_PREDICT]\n",
        "matrix = metrics.confusion_matrix(y_test, Y_PREDICT)\n",
        "recall = matrix[0][0]/(matrix[0][0]+matrix[1][0])\n",
        "precision = matrix[0][0]/(matrix[0][0]+matrix[0][1])\n",
        "print('Test loss:', score[0])\n",
        "print('Test total recall:', recall)\n",
        "print('Test total precision:', precision)\n",
        "print('Test total accuracy :', score[1])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test loss: 0.12484388020303515\n",
            "Test total recall: 0.813953488372093\n",
            "Test total precision: 0.8433734939759037\n",
            "Test total accuracy : 0.8388888915379842\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I0TClw2FSDUB",
        "colab_type": "text"
      },
      "source": [
        "# 2) Logistic Regression From Scratch"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EMNrrKhLSDUF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Your implementation\n",
        "def initialize_with_zeros(dim):\n",
        "    \n",
        "    #This function creates a vector of zeros of shape (dim, 1) for w and initializes b to 0.\n",
        "    \n",
        "    w = np.zeros((dim,1))\n",
        "    b = 0\n",
        "    assert(w.shape == (dim, 1))\n",
        "    assert(isinstance(b, float) or isinstance(b, int))\n",
        "    \n",
        "    return w, b\n",
        "    #w: initialized vector of shape (dim, 1)\n",
        "    #b: initialized scalar (corresponds to the bias)\n",
        "\n",
        "def sigmoid(z):\n",
        "    return 1 / (1 + np.exp(-z))\n",
        "\n",
        "def cost_function(y, a):\n",
        "    return -np.mean(y*np.log(a) + (1-y)*np.log(1-a))\n",
        "\n",
        "def optimize(w, b, X, Y, num_epochs, lr, print_cost = False):\n",
        "    costs = []\n",
        "    for epoch in range(num_epochs):\n",
        "    \n",
        "        # FORWARD PROPAGATION\n",
        "        m = X.shape[1] #mitone hazf she\n",
        "        Z = np.dot(w.T, X) + b #linear mult\n",
        "        A = sigmoid(Z)                                 \n",
        "        cost = cost_function(Y, A)  \n",
        "        cost = np.squeeze(cost)\n",
        "        assert(cost.shape == ())\n",
        "        back_require = { 'A': A }\n",
        "\n",
        "        # BACKWARD PROPAGATION\n",
        "        m = X.shape[1]\n",
        "        A = back_require['A']\n",
        "        dw = (1/m) * np.dot(X,(A-Y).T)\n",
        "        db = (1/m) * np.sum(A - Y)\n",
        "        assert(dw.shape == w.shape)\n",
        "        assert(db.dtype == float)\n",
        "        grads = {\"dw\": dw,\"db\": db}\n",
        "\n",
        "\n",
        "        dw = grads[\"dw\"] #weight hamon\n",
        "        db = grads[\"db\"] #bias ro sefr dar nazar gereftim\n",
        "\n",
        "        w -= lr*dw\n",
        "        b -= lr*db\n",
        "\n",
        "        # Record the costs\n",
        "        costs.append(cost)\n",
        "\n",
        "        # Print the cost every 100 training iterations\n",
        "        if print_cost and epoch % 100 == 0:\n",
        "            print (\"Cost after iteration %i: %f\" %(epoch, cost))\n",
        "    \n",
        "    params = {\"w\": w, \"b\": b}\n",
        "    grads = {\"dw\": dw, \"db\": db}\n",
        "    \n",
        "    return params, grads, costs"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ib55C8EvSDUO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train = (np.asarray(X)).T\n",
        "y_train = (np.asarray(y)).T\n",
        "\n",
        "w, b = initialize_with_zeros(x_train.shape[0])\n",
        "params, grads, costs = optimize(w, b, x_train , y_train , num_epochs= 10000, lr = 1, print_cost = False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7RlojzUdSDUW",
        "colab_type": "text"
      },
      "source": [
        "# Prediction and Evaluation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A6LeHKkmSDUa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Your code\n",
        "def predict(w, b, X):\n",
        "    \n",
        "    m = X.shape[1]\n",
        "    Y_prediction = np.zeros((1,m))\n",
        "    \n",
        "    Z = np.dot(w.T, X) + b\n",
        "    A = sigmoid(Z)\n",
        "    \n",
        "    for i in range(m):\n",
        "        Y_prediction[0][i] = 1 if A[0][i] > .5 else 0\n",
        "    \n",
        "    assert(Y_prediction.shape == (1, m))\n",
        "    \n",
        "    return Y_prediction\n",
        "\n",
        "x_test = (np.asarray(X_test)).T\n",
        "y_test = (np.asarray(y_test))\n",
        "\n",
        "prediction = predict(params['w'], params['b'], np.asarray(x_test))\n",
        "preds = prediction[0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ra_OsKS5SDUg",
        "colab_type": "text"
      },
      "source": [
        "# Confusion Matrix From Scratch\n",
        "\\\n",
        "$$\\text{Accuracy} = \\frac{TP + TN}{TP + FP + FN + TN}$$\n",
        "When it's actually yes, how often does it predict yes\n",
        "$$\\text{Precision} = \\frac{TP}{TP + FP}$$\n",
        "When it predicts yes, how often is it correct\n",
        "$$\\text{Recall} = \\frac{TP}{TP + FN}$$\n",
        "When it's actually yes, how often does it predict yes\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZUfOlfrDSDUi",
        "colab_type": "code",
        "outputId": "49ed95f1-0173-4392-c0a2-c18115cbe9be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 75
        }
      },
      "source": [
        "def ConfusionMatrix(preds, Y_test):\n",
        "    cm = [[0,0],[0,0]]\n",
        "    for i in range(0, len(preds)):\n",
        "        if preds[i] == Y_test[i]:\n",
        "            if preds[i] == 0:\n",
        "                cm[1][0] += 1\n",
        "            else:\n",
        "                cm[0][0] += 1\n",
        "        elif preds[i] == 0 and Y_test[i] == 1:\n",
        "            cm[0][1] += 1\n",
        "        else:\n",
        "            cm[1][1] += 1\n",
        "    return cm\n",
        "\n",
        "matrix = ConfusionMatrix(preds, y_test)\n",
        "accuracy = (matrix[0][0]+matrix[1][0])/(matrix[0][0]+matrix[0][1]+matrix[1][0]+matrix[1][1])\n",
        "print('Accuracy: ' + str(accuracy))\n",
        "precision = matrix[0][0]/(matrix[0][0]+matrix[0][1])\n",
        "print('Precision: ' + str(precision))\n",
        "recall = matrix[0][0]/(matrix[0][0]+matrix[1][0])\n",
        "print('Recall: ' + str(recall))\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy: 0.7722222222222223\n",
            "Precision: 0.845360824742268\n",
            "Recall: 0.5899280575539568\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}