Linear_Regression_Model_Updated.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Linear Regression Model.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyP4/XFt7LbRL2WM+nsehP0m",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ayushbisht2001/MACHINE-LEARNING-WITH-TF/blob/main/Linear_Regression_Model_Updated.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6Ooe0EuB1P6g"
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd"
      ],
      "execution_count": 193,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W_pjcdul3Qaj",
        "outputId": "248123b8-4d9a-40ab-ae41-67100650c719"
      },
      "source": [
        "tf.compat.v1.disable_eager_execution()\n",
        "x_data = np.linspace(0.0,10.0,1000000)\n",
        "noise = np.random.randn(len(x_data))\n",
        "noise"
      ],
      "execution_count": 194,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 0.84685295, -0.48265851, -1.30771946, ..., -0.47679332,\n",
              "        2.62339029,  1.76588725])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 194
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNImiWL63xw2"
      },
      "source": [
        "y_true = (1*x_data) + 5 + noise"
      ],
      "execution_count": 195,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A6eBirXJ4U9k"
      },
      "source": [
        "x_df = pd.DataFrame(data = x_data,columns = ['X-Data'])"
      ],
      "execution_count": 196,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v65ZYgPs4dg0"
      },
      "source": [
        "y_df = pd.DataFrame(data =y_true,columns=['Y'])"
      ],
      "execution_count": 197,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        },
        "id": "lLMqKm624xhT",
        "outputId": "d7052ba2-8ed8-4433-d78d-1cab5b5f8c1d"
      },
      "source": [
        "y_df.head()\n"
      ],
      "execution_count": 198,
      "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>Y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5.846853</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4.517351</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3.692301</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4.278523</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2.758584</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          Y\n",
              "0  5.846853\n",
              "1  4.517351\n",
              "2  3.692301\n",
              "3  4.278523\n",
              "4  2.758584"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 198
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eEqRVobs4zSr"
      },
      "source": [
        "my_data = pd.concat([x_df,y_df],axis=1)"
      ],
      "execution_count": 199,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        },
        "id": "XRnRNrU_5cDh",
        "outputId": "678c3be1-77b3-4231-971f-5a40d0079de3"
      },
      "source": [
        "my_data.head()"
      ],
      "execution_count": 200,
      "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>X-Data</th>\n",
              "      <th>Y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.00000</td>\n",
              "      <td>5.846853</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.00001</td>\n",
              "      <td>4.517351</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.00002</td>\n",
              "      <td>3.692301</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.00003</td>\n",
              "      <td>4.278523</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.00004</td>\n",
              "      <td>2.758584</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    X-Data         Y\n",
              "0  0.00000  5.846853\n",
              "1  0.00001  4.517351\n",
              "2  0.00002  3.692301\n",
              "3  0.00003  4.278523\n",
              "4  0.00004  2.758584"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 200
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "FwKb51Q25fGR",
        "outputId": "cb406908-4019-469e-d8e8-22d2852f4524"
      },
      "source": [
        "my_data.sample(n=250).plot(kind='scatter',x=\"X-Data\",y=\"Y\")"
      ],
      "execution_count": 201,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2a6b60940>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 201
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HEYvf2oW6Kt3"
      },
      "source": [
        "batch_size = 8"
      ],
      "execution_count": 202,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EjyzgUeSzg9K"
      },
      "source": [
        "* In short, you use tf.Variable for trainable variables such as weights (W) *and biases (B) for your model.\r\n",
        "\r\n",
        "* tf.placeholder is used to feed actual training examples."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rrrPG6NacFsD"
      },
      "source": [
        "m = tf.Variable(0.5)\n",
        "b = tf.Variable(1.0)"
      ],
      "execution_count": 203,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PhMEp1mBcJkZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7b8b11d2-f174-4622-96a7-b8a9f354957e"
      },
      "source": [
        "xph = tf.compat.v1.placeholder(tf.float32,[batch_size])\n",
        "yph = tf.compat.v1.placeholder(tf.float32 , [batch_size])\n",
        "\n",
        "xph"
      ],
      "execution_count": 204,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor 'Placeholder_14:0' shape=(8,) dtype=float32>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 204
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bdpP_q_gkyl4"
      },
      "source": [
        "y_model = m*xph + b"
      ],
      "execution_count": 205,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kuRsXo8-1dmN"
      },
      "source": [
        "* tk.reduce_sum() :  https://stackoverflow.com/questions/47157692/how-does-reduce-sum-work-in-tensorflow"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I8_D8RD9lm6U"
      },
      "source": [
        "error =  tf.reduce_sum(tf.square(yph-y_model))\r\n"
      ],
      "execution_count": 206,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fwqBFaCTlz_J"
      },
      "source": [
        "optimizer = tf.compat.v1.train.GradientDescentOptimizer(learning_rate = 0.001)\n",
        "train = optimizer.minimize(error)"
      ],
      "execution_count": 207,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AT_ST7bDmEDh"
      },
      "source": [
        "init = tf.compat.v1.global_variables_initializer()\n"
      ],
      "execution_count": 208,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AcaQ8Qq24rFb"
      },
      "source": [
        "* numpy.random.randint(low, high=None, size=None, dtype=int)\r\n",
        "  * Return random integers from low (inclusive) to high (exclusive).\r\n",
        "\r\n",
        "  * Return random integers from the “discrete uniform” distribution of the specified dtype in the “half-open” interval [low, high). If high is None (the default), then results are from [0, low).\r\n",
        "  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wjZDJOasnAJm"
      },
      "source": [
        "with tf.compat.v1.Session() as sess:\n",
        "  sess.run(init)\n",
        "  epochs = 100\n",
        "  for i in range(epochs):\n",
        "    rand_ind = np.random.randint(low=len(x_data),size = batch_size)    \n",
        "    feed = {xph : x_data[rand_ind],yph : y_true[rand_ind]}\n",
        "    sess.run(train,feed_dict = feed)\n",
        "  model_m,model_b = sess.run([m,b])"
      ],
      "execution_count": 209,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i6yvWGtPokbB"
      },
      "source": [
        "y_hat = x_data*model_m + model_b"
      ],
      "execution_count": 210,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LYxtzbMkp3uJ",
        "outputId": "44d93295-f81a-4bc5-a618-b7d06720f8ba"
      },
      "source": [
        "model_b"
      ],
      "execution_count": 211,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2.382064"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 211
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "f0yOQHvqpW6d",
        "outputId": "4d6039c4-d533-4079-aef0-794ffad81f93"
      },
      "source": [
        "my_data.sample(n=1000000).plot(kind='scatter',x='X-Data',y='Y')\n",
        "plt.plot(x_data,y_hat,'r')"
      ],
      "execution_count": 212,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fd2a607d2b0>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 212
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dWSYJX0oUpA2"
      },
      "source": [
        "Till now, We have manually created our machine learning model . \r\n",
        "\r\n",
        "Now the same thing can be done using APIs, which is implemented below...."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1nyfbNU7Qz-Y",
        "outputId": "abd18ef6-2f56-482c-c618-b268ef828eac"
      },
      "source": [
        "feat_cols = [tf.feature_column.numeric_column('x',shape=[1])]\r\n",
        "print(feat_cols)\r\n",
        "\r\n",
        "\r\n",
        "\r\n",
        " "
      ],
      "execution_count": 213,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[NumericColumn(key='x', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None)]\n",
            "INFO:tensorflow:Using default config.\n",
            "WARNING:tensorflow:Using temporary folder as model directory: /tmp/tmpzifk4h6x\n",
            "INFO:tensorflow:Using config: {'_model_dir': '/tmp/tmpzifk4h6x', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': allow_soft_placement: true\n",
            "graph_options {\n",
            "  rewrite_options {\n",
            "    meta_optimizer_iterations: ONE\n",
            "  }\n",
            "}\n",
            ", '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_experimental_max_worker_delay_secs': None, '_session_creation_timeout_secs': 7200, '_checkpoint_save_graph_def': True, '_service': None, '_cluster_spec': ClusterSpec({}), '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pd2q2kbE2-U0"
      },
      "source": [
        "## tf.estimator.LinearRegressor(feature_columns=feat_cols)\r\n",
        "\r\n",
        "* It basically do all the stuff by themselve, such that, taking some value for m and b and then apply function with Gradient optimizer. Basically we are providing the x and y value, and it perform linear regression for these given values to get the predicted value for y."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Akg7Hpsm2_Rb"
      },
      "source": [
        "# estimator is basically use to call the different form of model.\r\n",
        "estimator = tf.estimator.LinearRegressor(feature_columns = feat_cols)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2HV-tynuRI7h",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "48b60e6d-0ae7-4665-f545-eee2f649c148"
      },
      "source": [
        "from sklearn.model_selection import train_test_split\r\n",
        "\r\n",
        "# train_test_split is used to split the data into training and test set with 30% of the total data in test set and remainning are in training set\r\n",
        "# https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html\r\n",
        "\r\n",
        "\r\n",
        "x_train , x_eval,y_train,y_eval = train_test_split(x_data,y_true,test_size =0.3,random_state=0)\r\n",
        "\r\n",
        "\r\n",
        "input_func = tf.compat.v1.estimator.inputs.numpy_input_fn({'x':x_train},y_train,batch_size=4,num_epochs=None,shuffle = True)\r\n",
        "print(input_func)\r\n",
        "train_input_func = tf.compat.v1.estimator.inputs.numpy_input_fn({'x':x_train},y_train,batch_size=4,num_epochs=100,shuffle = False)\r\n",
        "\r\n",
        "eval_input_func = tf.compat.v1.estimator.inputs.numpy_input_fn({'x':x_train},y_train,batch_size=4,num_epochs=100,shuffle = False)"
      ],
      "execution_count": 214,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<function numpy_input_fn.<locals>.input_fn at 0x7fd2a5635620>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eL-cAlPARNGZ",
        "outputId": "3d54177a-94b0-4812-d638-78af7698709a"
      },
      "source": [
        "estimator.train(input_fn = input_func,steps = 1000)\r\n",
        "\r\n",
        "\r\n"
      ],
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Calling model_fn.\n",
            "INFO:tensorflow:Done calling model_fn.\n",
            "INFO:tensorflow:Create CheckpointSaverHook.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/base_layer_v1.py:1727: UserWarning: `layer.add_variable` is deprecated and will be removed in a future version. Please use `layer.add_weight` method instead.\n",
            "  warnings.warn('`layer.add_variable` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Graph was finalized.\n",
            "INFO:tensorflow:Running local_init_op.\n",
            "INFO:tensorflow:Done running local_init_op.\n",
            "INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 0...\n",
            "INFO:tensorflow:Saving checkpoints for 0 into /tmp/tmpzifk4h6x/model.ckpt.\n",
            "INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 0...\n",
            "INFO:tensorflow:loss = 109.86784, step = 0\n",
            "INFO:tensorflow:global_step/sec: 752.006\n",
            "INFO:tensorflow:loss = 5.6724787, step = 100 (0.135 sec)\n",
            "INFO:tensorflow:global_step/sec: 762.914\n",
            "INFO:tensorflow:loss = 1.2566943, step = 201 (0.132 sec)\n",
            "INFO:tensorflow:global_step/sec: 783.541\n",
            "INFO:tensorflow:loss = 2.3904824, step = 300 (0.128 sec)\n",
            "INFO:tensorflow:global_step/sec: 808.849\n",
            "INFO:tensorflow:loss = 0.9993079, step = 400 (0.125 sec)\n",
            "INFO:tensorflow:global_step/sec: 796.845\n",
            "INFO:tensorflow:loss = 2.5215938, step = 500 (0.132 sec)\n",
            "INFO:tensorflow:global_step/sec: 682.618\n",
            "INFO:tensorflow:loss = 0.84576976, step = 600 (0.137 sec)\n",
            "INFO:tensorflow:global_step/sec: 842.358\n",
            "INFO:tensorflow:loss = 0.10960514, step = 700 (0.119 sec)\n",
            "INFO:tensorflow:global_step/sec: 754.814\n",
            "INFO:tensorflow:loss = 2.113506, step = 801 (0.133 sec)\n",
            "INFO:tensorflow:global_step/sec: 812.273\n",
            "INFO:tensorflow:loss = 0.25273216, step = 900 (0.123 sec)\n",
            "INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 1000...\n",
            "INFO:tensorflow:Saving checkpoints for 1000 into /tmp/tmpzifk4h6x/model.ckpt.\n",
            "INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 1000...\n",
            "INFO:tensorflow:Loss for final step: 0.20968705.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow_estimator.python.estimator.canned.linear.LinearRegressorV2 at 0x7fd2a607d0f0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 215
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Oqmr5WPE1Yk2"
      },
      "source": [
        "## Getting performance metrics , such that: average_loss, mean, etc etc...\r\n",
        "* They are just to check the performance of model.If evaluation is not done,our model still give the same output."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ABoiOjeC1Xcx"
      },
      "source": [
        "train_metrics = estimator.evaluate(input_fn=train_input_func,steps = 1000)\r\n",
        "\r\n",
        "eval_metrics = estimator.evaluate(input_fn=eval_input_func ,steps = 1000)                                                                                                                                          \r\n",
        "\r\n",
        "input_fu_predict = tf.compat.v1.estimator.inputs.numpy_input_fn({'x':np.linspace(0,10,10)},shuffle = False)"
      ],
      "execution_count": 216,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "82IoZIkNRR7i",
        "outputId": "823b62f4-026e-4827-faf2-f0843f7440c7"
      },
      "source": [
        "Predictions = [x['predictions'] for x in estimator.predict(input_fn=input_fu_predict)]"
      ],
      "execution_count": 217,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Calling model_fn.\n",
            "INFO:tensorflow:Done calling model_fn.\n",
            "INFO:tensorflow:Graph was finalized.\n",
            "INFO:tensorflow:Restoring parameters from /tmp/tmpzifk4h6x/model.ckpt-1000\n",
            "INFO:tensorflow:Running local_init_op.\n",
            "INFO:tensorflow:Done running local_init_op.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/base_layer_v1.py:1727: UserWarning: `layer.add_variable` is deprecated and will be removed in a future version. Please use `layer.add_weight` method instead.\n",
            "  warnings.warn('`layer.add_variable` is deprecated and '\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tsl1TgBmYUDc",
        "outputId": "26e84227-6f1c-4790-97a2-9879b1e075cd"
      },
      "source": [
        "Predictions"
      ],
      "execution_count": 218,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([4.0794535], dtype=float32),\n",
              " array([5.3695397], dtype=float32),\n",
              " array([6.659626], dtype=float32),\n",
              " array([7.9497128], dtype=float32),\n",
              " array([9.2397995], dtype=float32),\n",
              " array([10.529885], dtype=float32),\n",
              " array([11.819971], dtype=float32),\n",
              " array([13.110058], dtype=float32),\n",
              " array([14.400145], dtype=float32),\n",
              " array([15.69023], dtype=float32)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 218
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "x1nZWiTvYgO6",
        "outputId": "bebafa1f-3721-497d-cd10-99d0ae3312b9"
      },
      "source": [
        "my_data.sample(n=200).plot(kind='scatter',x='X-Data',y='Y')\r\n",
        "plt.plot(np.linspace(0,10,10),Predictions,'r')"
      ],
      "execution_count": 219,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fd2a4b469e8>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 219
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zJo9S5jtZWH3"
      },
      "source": [
        ""
      ],
      "execution_count": 219,
      "outputs": []
    }
  ]
}