dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Download the MNIST dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "import utils\n",
    "from tensorflow.contrib.learn.python.learn.datasets import mnist\n",
    "import tensorflow as tf\n",
    "\n",
    "data_sets = mnist.read_data_sets('data', dtype=tf.uint8, reshape=False, validation_size=5000)\n",
    "\n",
    "utils.convert_to(data_sets.train, 'train', 'data')\n",
    "utils.convert_to(data_sets.validation, 'validation', 'data')\n",
    "utils.convert_to(data_sets.test, 'test', 'data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<TFRecordDataset shapes: (), types: tf.string>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import os\n",
    "tfrecord_dir = 'data'\n",
    "type=\"test\"\n",
    "tfrecord_file = [os.path.join(tfrecord_dir,f) for f in os.listdir(tfrecord_dir) if type in f and \".tfrecord\" in f]\n",
    "dataset = tf.data.TFRecordDataset([tfrecord_file])\n",
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<RepeatDataset shapes: (), types: tf.string>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = dataset.repeat()\n",
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<PrefetchDataset shapes: ((?, 784), (?,)), types: (tf.float32, tf.int32)>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def parse_record(serialized_example):\n",
    "    features = tf.parse_single_example(\n",
    "        serialized_example,\n",
    "        features={\n",
    "            'image_raw': tf.FixedLenFeature([], tf.string),\n",
    "            'label': tf.FixedLenFeature([], tf.int64),\n",
    "        })\n",
    "\n",
    "    image = tf.decode_raw(features['image_raw'], tf.uint8)\n",
    "    image.set_shape([784])\n",
    "    image = tf.cast(image, tf.float32) * (1. / 255)\n",
    "    label = tf.cast(features['label'], tf.int32)\n",
    "\n",
    "    return image, label\n",
    "\n",
    "\n",
    "shuffle_buffer_size = 5000\n",
    "batch_size = 128\n",
    "num_parallel_batches=1\n",
    "prefetch_buffer_size = 5000\n",
    "\n",
    "\n",
    "dataset = dataset.shuffle(buffer_size=shuffle_buffer_size)\n",
    "\n",
    "\n",
    "dataset = dataset.apply(tf.contrib.data.map_and_batch(\n",
    "    map_func=parse_record, batch_size=batch_size, num_parallel_batches=num_parallel_batches))\n",
    "\n",
    "dataset = dataset.prefetch(buffer_size=prefetch_buffer_size)\n",
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(array([[0., 0., 0., ..., 0., 0., 0.],\n",
      "       [0., 0., 0., ..., 0., 0., 0.],\n",
      "       [0., 0., 0., ..., 0., 0., 0.],\n",
      "       ...,\n",
      "       [0., 0., 0., ..., 0., 0., 0.],\n",
      "       [0., 0., 0., ..., 0., 0., 0.],\n",
      "       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32), array([6, 7, 1, 8, 4, 1, 3, 9, 8, 4, 9, 4, 6, 9, 6, 9, 1, 0, 7, 1, 4, 2,\n",
      "       9, 6, 1, 2, 1, 1, 2, 7, 1, 2, 1, 8, 4, 7, 9, 4, 3, 5, 8, 4, 4, 4,\n",
      "       1, 5, 7, 8, 3, 1, 5, 5, 2, 6, 5, 1, 6, 9, 8, 2, 6, 0, 3, 6, 6, 1,\n",
      "       3, 5, 3, 9, 9, 1, 8, 9, 1, 8, 8, 2, 5, 1, 0, 7, 0, 4, 5, 5, 9, 0,\n",
      "       2, 7, 3, 0, 2, 0, 8, 7, 7, 6, 6, 8, 3, 5, 6, 7, 6, 4, 0, 8, 9, 5,\n",
      "       9, 9, 9, 4, 4, 5, 8, 3, 4, 8, 7, 1, 4, 7, 3, 7, 5, 1], dtype=int32))\n"
     ]
    }
   ],
   "source": [
    "iterator = dataset.make_one_shot_iterator()\n",
    "next_element = iterator.get_next()\n",
    "\n",
    "sess = tf.Session()\n",
    "print(sess.run(next_element))"
   ]
  }
 ],
 "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": "3.6.7"
  },
  "notice": "Copyright 2017 Amazon.com, Inc. or its affiliates. All Rights Reserved.  Licensed under the Apache License, Version 2.0 (the \"License\"). You may not use this file except in compliance with the License. A copy of the License is located at http://aws.amazon.com/apache2.0/ or in the \"license\" file accompanying this file. This file is distributed on an \"AS IS\" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License."
 },
 "nbformat": 4,
 "nbformat_minor": 2
}