TensorflowTest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "view-in-github"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/timurista/data-analysis/blob/master/TensorflowTest.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 513
    },
    "colab_type": "code",
    "id": "1Q8yZ-dfMaRX",
    "outputId": "eaa472e6-52d1-44cc-8a56-7aa3df781795"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow==2.0.0-beta1\n",
      "\u001b[31m  Could not find a version that satisfies the requirement tensorflow==2.0.0-beta1 (from versions: )\u001b[0m\n",
      "\u001b[31mNo matching distribution found for tensorflow==2.0.0-beta1\u001b[0m\n",
      "Requirement already satisfied: matplotlib in /anaconda3/lib/python3.7/site-packages (3.0.3)\n",
      "Requirement already satisfied: numpy>=1.10.0 in /anaconda3/lib/python3.7/site-packages (from matplotlib) (1.16.2)\n",
      "Requirement already satisfied: cycler>=0.10 in /anaconda3/lib/python3.7/site-packages (from matplotlib) (0.10.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /anaconda3/lib/python3.7/site-packages (from matplotlib) (1.0.1)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /anaconda3/lib/python3.7/site-packages (from matplotlib) (2.3.1)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in /anaconda3/lib/python3.7/site-packages (from matplotlib) (2.8.0)\n",
      "Requirement already satisfied: six in /anaconda3/lib/python3.7/site-packages (from cycler>=0.10->matplotlib) (1.12.0)\n",
      "Requirement already satisfied: setuptools in /anaconda3/lib/python3.7/site-packages (from kiwisolver>=1.0.1->matplotlib) (40.8.0)\n"
     ]
    }
   ],
   "source": [
    "# Install TensorFlow\n",
    "!pip install tensorflow==2.0.0-beta1\n",
    "!pip install matplotlib\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1wOmmpKVFzXU"
   },
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "T-Y02p7qNfNK",
    "outputId": "4f4012c8-e568-4cbd-db12-bc30d06a6b1c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0.0-beta1\n",
      "1.16.4\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "print(tf.__version__)\n",
    "print(np.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "CAELGpRmM1NU"
   },
   "source": [
    "# Initial Start"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "lbgZ-_RqMa3X"
   },
   "source": [
    "Python 3 tensorflow example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "a1txH-CFM6ur",
    "outputId": "7d54c2cc-8bcb-49d0-99a6-7ec14a4a8e4e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000, 2) (1000, 1)\n"
     ]
    }
   ],
   "source": [
    "observations = 1000\n",
    "xs = np.random.uniform(low=-10,high=10,size=(observations, 1))\n",
    "zs = np.random.uniform(-10,10,(observations, 1))\n",
    "\n",
    "inputs = np.column_stack((xs,zs))\n",
    "\n",
    "noise = np.random.uniform(-1,1,(observations, 1))\n",
    "\n",
    "targets = 2*xs - 3*zs + 5 + noise\n",
    "\n",
    "print(inputs.shape, noise.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "5Qn3u6mLM8Vr"
   },
   "outputs": [],
   "source": [
    "np.savez('TF_intro', inputs=inputs, targets=targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ihEkO7psGWq9"
   },
   "outputs": [],
   "source": [
    "training_data = np.load('TF_intro.npz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "2hLFVXonGqIJ",
    "outputId": "db65b1ab-f5dc-4858-cbf0-f94fff36fde0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1000 samples\n",
      "Epoch 1/100\n",
      "1000/1000 - 0s - loss: 22.8170\n",
      "Epoch 2/100\n",
      "1000/1000 - 0s - loss: 1.4698\n",
      "Epoch 3/100\n",
      "1000/1000 - 0s - loss: 0.4385\n",
      "Epoch 4/100\n",
      "1000/1000 - 0s - loss: 0.3748\n",
      "Epoch 5/100\n",
      "1000/1000 - 0s - loss: 0.3960\n",
      "Epoch 6/100\n",
      "1000/1000 - 0s - loss: 0.4134\n",
      "Epoch 7/100\n",
      "1000/1000 - 0s - loss: 0.4142\n",
      "Epoch 8/100\n",
      "1000/1000 - 0s - loss: 0.3638\n",
      "Epoch 9/100\n",
      "1000/1000 - 0s - loss: 0.4030\n",
      "Epoch 10/100\n",
      "1000/1000 - 0s - loss: 0.4956\n",
      "Epoch 11/100\n",
      "1000/1000 - 0s - loss: 0.4022\n",
      "Epoch 12/100\n",
      "1000/1000 - 0s - loss: 0.3674\n",
      "Epoch 13/100\n",
      "1000/1000 - 0s - loss: 0.3931\n",
      "Epoch 14/100\n",
      "1000/1000 - 0s - loss: 0.4370\n",
      "Epoch 15/100\n",
      "1000/1000 - 0s - loss: 0.3896\n",
      "Epoch 16/100\n",
      "1000/1000 - 0s - loss: 0.3812\n",
      "Epoch 17/100\n",
      "1000/1000 - 0s - loss: 0.4129\n",
      "Epoch 18/100\n",
      "1000/1000 - 0s - loss: 0.4536\n",
      "Epoch 19/100\n",
      "1000/1000 - 0s - loss: 0.3819\n",
      "Epoch 20/100\n",
      "1000/1000 - 0s - loss: 0.3883\n",
      "Epoch 21/100\n",
      "1000/1000 - 0s - loss: 0.4072\n",
      "Epoch 22/100\n",
      "1000/1000 - 0s - loss: 0.4200\n",
      "Epoch 23/100\n",
      "1000/1000 - 0s - loss: 0.3792\n",
      "Epoch 24/100\n",
      "1000/1000 - 0s - loss: 0.3932\n",
      "Epoch 25/100\n",
      "1000/1000 - 0s - loss: 0.3815\n",
      "Epoch 26/100\n",
      "1000/1000 - 0s - loss: 0.3735\n",
      "Epoch 27/100\n",
      "1000/1000 - 0s - loss: 0.3645\n",
      "Epoch 28/100\n",
      "1000/1000 - 0s - loss: 0.4044\n",
      "Epoch 29/100\n",
      "1000/1000 - 0s - loss: 0.3588\n",
      "Epoch 30/100\n",
      "1000/1000 - 0s - loss: 0.3532\n",
      "Epoch 31/100\n",
      "1000/1000 - 0s - loss: 0.3674\n",
      "Epoch 32/100\n",
      "1000/1000 - 0s - loss: 0.3719\n",
      "Epoch 33/100\n",
      "1000/1000 - 0s - loss: 0.4452\n",
      "Epoch 34/100\n",
      "1000/1000 - 0s - loss: 0.3529\n",
      "Epoch 35/100\n",
      "1000/1000 - 0s - loss: 0.3624\n",
      "Epoch 36/100\n",
      "1000/1000 - 0s - loss: 0.4141\n",
      "Epoch 37/100\n",
      "1000/1000 - 0s - loss: 0.4138\n",
      "Epoch 38/100\n",
      "1000/1000 - 0s - loss: 0.3804\n",
      "Epoch 39/100\n",
      "1000/1000 - 0s - loss: 0.4507\n",
      "Epoch 40/100\n",
      "1000/1000 - 0s - loss: 0.4427\n",
      "Epoch 41/100\n",
      "1000/1000 - 0s - loss: 0.3688\n",
      "Epoch 42/100\n",
      "1000/1000 - 0s - loss: 0.4198\n",
      "Epoch 43/100\n",
      "1000/1000 - 0s - loss: 0.3937\n",
      "Epoch 44/100\n",
      "1000/1000 - 0s - loss: 0.4112\n",
      "Epoch 45/100\n",
      "1000/1000 - 0s - loss: 0.3803\n",
      "Epoch 46/100\n",
      "1000/1000 - 0s - loss: 0.3725\n",
      "Epoch 47/100\n",
      "1000/1000 - 0s - loss: 0.3798\n",
      "Epoch 48/100\n",
      "1000/1000 - 0s - loss: 0.3739\n",
      "Epoch 49/100\n",
      "1000/1000 - 0s - loss: 0.3896\n",
      "Epoch 50/100\n",
      "1000/1000 - 0s - loss: 0.3829\n",
      "Epoch 51/100\n",
      "1000/1000 - 0s - loss: 0.3926\n",
      "Epoch 52/100\n",
      "1000/1000 - 0s - loss: 0.3800\n",
      "Epoch 53/100\n",
      "1000/1000 - 0s - loss: 0.3993\n",
      "Epoch 54/100\n",
      "1000/1000 - 0s - loss: 0.3986\n",
      "Epoch 55/100\n",
      "1000/1000 - 0s - loss: 0.4128\n",
      "Epoch 56/100\n",
      "1000/1000 - 0s - loss: 0.4060\n",
      "Epoch 57/100\n",
      "1000/1000 - 0s - loss: 0.4079\n",
      "Epoch 58/100\n",
      "1000/1000 - 0s - loss: 0.4210\n",
      "Epoch 59/100\n",
      "1000/1000 - 0s - loss: 0.3643\n",
      "Epoch 60/100\n",
      "1000/1000 - 0s - loss: 0.4180\n",
      "Epoch 61/100\n",
      "1000/1000 - 0s - loss: 0.3784\n",
      "Epoch 62/100\n",
      "1000/1000 - 0s - loss: 0.3853\n",
      "Epoch 63/100\n",
      "1000/1000 - 0s - loss: 0.3818\n",
      "Epoch 64/100\n",
      "1000/1000 - 0s - loss: 0.4125\n",
      "Epoch 65/100\n",
      "1000/1000 - 0s - loss: 0.4080\n",
      "Epoch 66/100\n",
      "1000/1000 - 0s - loss: 0.3878\n",
      "Epoch 67/100\n",
      "1000/1000 - 0s - loss: 0.3546\n",
      "Epoch 68/100\n",
      "1000/1000 - 0s - loss: 0.3666\n",
      "Epoch 69/100\n",
      "1000/1000 - 0s - loss: 0.4044\n",
      "Epoch 70/100\n",
      "1000/1000 - 0s - loss: 0.3840\n",
      "Epoch 71/100\n",
      "1000/1000 - 0s - loss: 0.4658\n",
      "Epoch 72/100\n",
      "1000/1000 - 0s - loss: 0.3780\n",
      "Epoch 73/100\n",
      "1000/1000 - 0s - loss: 0.3793\n",
      "Epoch 74/100\n",
      "1000/1000 - 0s - loss: 0.4025\n",
      "Epoch 75/100\n",
      "1000/1000 - 0s - loss: 0.4185\n",
      "Epoch 76/100\n",
      "1000/1000 - 0s - loss: 0.3910\n",
      "Epoch 77/100\n",
      "1000/1000 - 0s - loss: 0.3819\n",
      "Epoch 78/100\n",
      "1000/1000 - 0s - loss: 0.3882\n",
      "Epoch 79/100\n",
      "1000/1000 - 0s - loss: 0.3796\n",
      "Epoch 80/100\n",
      "1000/1000 - 0s - loss: 0.3769\n",
      "Epoch 81/100\n",
      "1000/1000 - 0s - loss: 0.3945\n",
      "Epoch 82/100\n",
      "1000/1000 - 0s - loss: 0.3760\n",
      "Epoch 83/100\n",
      "1000/1000 - 0s - loss: 0.3847\n",
      "Epoch 84/100\n",
      "1000/1000 - 0s - loss: 0.4134\n",
      "Epoch 85/100\n",
      "1000/1000 - 0s - loss: 0.3823\n",
      "Epoch 86/100\n",
      "1000/1000 - 0s - loss: 0.4023\n",
      "Epoch 87/100\n",
      "1000/1000 - 0s - loss: 0.4031\n",
      "Epoch 88/100\n",
      "1000/1000 - 0s - loss: 0.3937\n",
      "Epoch 89/100\n",
      "1000/1000 - 0s - loss: 0.3869\n",
      "Epoch 90/100\n",
      "1000/1000 - 0s - loss: 0.3882\n",
      "Epoch 91/100\n",
      "1000/1000 - 0s - loss: 0.3903\n",
      "Epoch 92/100\n",
      "1000/1000 - 0s - loss: 0.3838\n",
      "Epoch 93/100\n",
      "1000/1000 - 0s - loss: 0.3892\n",
      "Epoch 94/100\n",
      "1000/1000 - 0s - loss: 0.3881\n",
      "Epoch 95/100\n",
      "1000/1000 - 0s - loss: 0.3853\n",
      "Epoch 96/100\n",
      "1000/1000 - 0s - loss: 0.3629\n",
      "Epoch 97/100\n",
      "1000/1000 - 0s - loss: 0.3666\n",
      "Epoch 98/100\n",
      "1000/1000 - 0s - loss: 0.3652\n",
      "Epoch 99/100\n",
      "1000/1000 - 0s - loss: 0.4228\n",
      "Epoch 100/100\n",
      "1000/1000 - 0s - loss: 0.3891\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7f5252aa4860>"
      ]
     },
     "execution_count": 27,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "input_size = 2\n",
    "output_size = 1\n",
    "\n",
    "model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Dense(output_size)\n",
    "])\n",
    "\n",
    "## cusotom optimizer\n",
    "custom_optimizer = tf.keras.optimizers.SGD(learning_rate=0.02)\n",
    "\n",
    "## sgd is stochastic gradient descent\n",
    "model.compile(custom_optimizer, loss=\"mean_squared_error\")\n",
    "model.fit(training_data['inputs'], training_data['targets'], epochs=100, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "jVNXHpZgICSk",
    "outputId": "6a6d6456-8e2a-4859-aeea-b6603aa4bc53"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[ 2.0229752],\n",
       "        [-2.9728978]], dtype=float32), array([5.0430207], dtype=float32)]"
      ]
     },
     "execution_count": 28,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.layers[0].get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "o3uOb0B3IaNa"
   },
   "outputs": [],
   "source": [
    "predicted = model.predict_on_batch(training_data['inputs']).round(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4j_UuehaIlnx"
   },
   "outputs": [],
   "source": [
    "actual = training_data['targets'].round(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Ned7skyOKEz7"
   },
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NeuaRdT3Is9q"
   },
   "outputs": [],
   "source": [
    "## Plot outputs vs targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 283
    },
    "colab_type": "code",
    "id": "bS0PlGs7JC5_",
    "outputId": "fc935480-db76-4379-98a0-4846e9dec695"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEKCAYAAAAMzhLIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.squeeze(predicted), np.squeeze(actual))\n",
    "plt.xlabel('output')\n",
    "plt.ylabel('targets')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4Rgej0HCJWTX"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "include_colab_link": true,
   "name": "TensorflowTest.ipynb",
   "provenance": [],
   "toc_visible": true,
   "version": "0.3.2"
  },
  "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}