bailulu
/
mindspore_yolo

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.2.0\n",
      "sys.version_info(major=3, minor=6, micro=9, releaselevel='final', serial=0)\n",
      "matplotlib 3.3.4\n",
      "numpy 1.19.5\n",
      "pandas 1.1.5\n",
      "sklearn 0.24.2\n",
      "tensorflow 2.2.0\n",
      "tensorflow.keras 2.3.0-tf\n"
     ]
    }
   ],
   "source": [
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import sklearn\n",
    "import pandas as pd\n",
    "import os\n",
    "import sys\n",
    "import time\n",
    "import tensorflow as tf\n",
    "\n",
    "from tensorflow import keras\n",
    "\n",
    "print(tf.__version__)\n",
    "print(sys.version_info)\n",
    "for module in mpl, np, pd, sklearn, tf, keras:\n",
    "    print(module.__name__, module.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# from_tensor_slices ,repeat,batch,interleave"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<TensorSliceDataset shapes: (), types: tf.int64>\n"
     ]
    }
   ],
   "source": [
    "#内存中构建dataset，初始化dataset，返回的像一个迭代器\n",
    "dataset = tf.data.Dataset.from_tensor_slices(np.arange(1,6).repeat(6))\n",
    "\n",
    "print(dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "#可以做遍历，每一个都是tensor\n",
    "for item in dataset:\n",
    "    print(item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "dataset1 = dataset.repeat(3)  #为了epoch服务的\n",
    "for item in dataset1:\n",
    "    print(item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<BatchDataset shapes: (None,), types: tf.int64>\n",
      "tf.Tensor([1 1 1 1 1 1 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([2 2 2 2 2 3 3], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 3 3 3 4 4 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([4 4 4 5 5 5 5], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 1 1 1 1 1], shape=(7,), dtype=int64)\n",
      "tf.Tensor([1 2 2 2 2 2 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 3 3 3 3 3 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([4 4 4 4 4 5 5], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 5 5 1 1 1], shape=(7,), dtype=int64)\n",
      "tf.Tensor([1 1 1 2 2 2 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([2 2 3 3 3 3 3], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 4 4 4 4 4 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 5 5 5 5], shape=(6,), dtype=int64)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "13"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1. repeat epoch 遍历一次数据集，就称为1次epoch\n",
    "# 2. get batch  取一部分数据集\n",
    "\n",
    "dataset = dataset.repeat(3).batch(7)\n",
    "print(dataset)\n",
    "i=0\n",
    "for item in dataset:\n",
    "    i=i+1\n",
    "    print(item)\n",
    "i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([1 1 1 1 1 1 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([2 2 2 2 2 3 3], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 3 3 3 4 4 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([4 4 4 5 5 5 5], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 1 1 1 1 1], shape=(7,), dtype=int64)\n",
      "tf.Tensor([1 2 2 2 2 2 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 3 3 3 3 3 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([4 4 4 4 4 5 5], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 5 5 1 1 1], shape=(7,), dtype=int64)\n",
      "tf.Tensor([1 1 1 2 2 2 2], shape=(7,), dtype=int64)\n",
      "tf.Tensor([2 2 3 3 3 3 3], shape=(7,), dtype=int64)\n",
      "tf.Tensor([3 4 4 4 4 4 4], shape=(7,), dtype=int64)\n",
      "tf.Tensor([5 5 5 5 5 5], shape=(6,), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "for i in dataset:\n",
    "    print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<InterleaveDataset shapes: (), types: tf.int64>\n",
      "--------------------------------------------------\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "90"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# interleave: 作用 https://zhuanlan.zhihu.com/p/97876668\n",
    "# case: 文件dataset -> 具体数据集\n",
    "\n",
    "dataset2 = dataset.interleave(\n",
    "    lambda v: tf.data.Dataset.from_tensor_slices(v), # map_fn，第一参数是回调函数\n",
    "    cycle_length = 5, # cycle_length,每一个cycle提取的个数\n",
    "    block_length = 6, # block_length\n",
    ")\n",
    "print(dataset2)\n",
    "print('-'*50)\n",
    "i=0\n",
    "for item in dataset2:\n",
    "    i=i+1\n",
    "    print(item)\n",
    "i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 下面的例子理解interleave更加简单"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "--------------------------------------------------\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "a = tf.data.Dataset.range(1,6)\n",
    "for i in a:\n",
    "    print(i)\n",
    "# b=a.repeat(6)\n",
    "# for i in b:\n",
    "#     print(i)\n",
    "print('-'*50)\n",
    "#cycle_length是使用了几个block以后，就要重复的  参数\n",
    "a1=a.interleave(lambda x: tf.data.Dataset.from_tensors(x).repeat(6),\n",
    "            cycle_length=3, block_length=4)\n",
    "for i in a1:\n",
    "    print(i)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 针对各种格式数据变为dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<TensorSliceDataset shapes: ((2,), ()), types: (tf.int64, tf.string)>\n",
      "tf.Tensor([1 2], shape=(2,), dtype=int64) tf.Tensor(b'cat', shape=(), dtype=string)\n",
      "tf.Tensor([3 4], shape=(2,), dtype=int64) tf.Tensor(b'dog', shape=(), dtype=string)\n",
      "tf.Tensor([5 6], shape=(2,), dtype=int64) tf.Tensor(b'fox', shape=(), dtype=string)\n"
     ]
    }
   ],
   "source": [
    "x = np.array([[1, 2], [3, 4], [5, 6]])\n",
    "y = np.array(['cat', 'dog', 'fox'])\n",
    "#输入的参数是元祖的情况下\n",
    "dataset3 = tf.data.Dataset.from_tensor_slices((x, y))\n",
    "print(dataset3)\n",
    "\n",
    "for item_x, item_y in dataset3:\n",
    "    print(item_x, item_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<TensorSliceDataset shapes: {feature1: (2,), label: ()}, types: {feature1: tf.int64, label: tf.string}>\n",
      "tf.Tensor([1 2], shape=(2,), dtype=int64) tf.Tensor(b'cat', shape=(), dtype=string)\n",
      "tf.Tensor([3 4], shape=(2,), dtype=int64) tf.Tensor(b'dog', shape=(), dtype=string)\n",
      "tf.Tensor([5 6], shape=(2,), dtype=int64) tf.Tensor(b'fox', shape=(), dtype=string)\n"
     ]
    }
   ],
   "source": [
    "#输入的参数是字典的情况下\n",
    "dataset4 = tf.data.Dataset.from_tensor_slices({\"feature1\": x,\n",
    "                                               \"label\": y})\n",
    "print(dataset4)\n",
    "for item in dataset4:\n",
    "    print(item[\"feature1\"], item[\"label\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<RangeDataset shapes: (), types: tf.int64>\n",
      "<InterleaveDataset shapes: (), types: tf.int64>\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(1, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(2, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(3, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(4, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n",
      "tf.Tensor(5, shape=(), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "a = tf.data.Dataset.range(1, 6)  # ==> [ 1, 2, 3, 4, 5 ]\n",
    "print(a)\n",
    "# NOTE: New lines indicate \"block\" boundaries.\n",
    "dataset2=a.interleave(lambda x: tf.data.Dataset.from_tensors(x).repeat(6),\n",
    "            cycle_length=2, block_length=4)  # ==> [1, 1, 1, 1,\n",
    "                                             #      2, 2, 2, 2,\n",
    "                                             #      1, 1,\n",
    "                                             #      2, 2,\n",
    "                                             #      3, 3, 3, 3,\n",
    "                                             #      4, 4, 4, 4,\n",
    "                                             #      3, 3,\n",
    "                                             #      4, 4,\n",
    "                                             #      5, 5, 5, 5,\n",
    "                                             #      5, 5]\n",
    "print(dataset2)\n",
    "for item in dataset2:\n",
    "    print(item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}