資料載入

目錄

• Outline
• keras.datasets
  • MNIST
  • CIFAR10/100
• tf.data.Dataset
  • .shuffle
  • .map
  • .batch
  • .repeat()
• For example

Outline

• keras.datasets

• tf.data.Dataset.from_tensor_slices
  • shuffle
  • map
  • batch
  • repeat

• will display Input Pipeline later（大資料集）

keras.datasets

• boston housing
  • Boston housing price regression dataset
• mnist/fashion mnist
  • MNIST/Fashion-MNIST dataset
• cifar10/100
  • small images classification dataset
• imdb
  • sentiment classification dataset

MNIST

import tensorflow as tf
from tensorflow import keras

# train: 60k | test: 10k
(x, y), (x_test, y_test) = keras.datasets.mnist.load_data()

x.shape

(60000, 28, 28)

y.shape

(60000,)

# 0純黑、255純白
x.min(), x.max(), x.mean()

(0, 255, 33.318421449829934)

x_test.shape, y_test.shape

((10000, 28, 28), (10000,))

y[:4]

array([5, 0, 4, 1], dtype=uint8)

# 0-9有10種分類結果
y_onehot = tf.one_hot(y, depth=10)
y_onehot[:2]

<tf.Tensor: id=13, shape=(2, 10), dtype=float32, numpy=
array([[0., 0., 0., 0., 0., 1., 0., 0., 0., 0.],
       [1., 0., 0., 0., 0., 0., 0., 0., 0., 0.]], dtype=float32)>

CIFAR10/100

• 10個大類中有100個小類

# train: 50k | test: 10k
(x, y), (x_test, y_test) = keras.datasets.cifar10.load_data()

x.shape,y.shape,x_test.shape,y_test.shape

x.min(),x.max()

y[:4]

tf.data.Dataset

• from_tensor_slices()

db =  tf.data.Dataset.from_tensor_slices(x_test)
next(iter(db)).shape

db = tf.data.Dataset.from_tensor_slices((x_test,y_test))
next(iter(db))[0].shape

.shuffle

• 打亂資料

db = tf.data.Dataset.from_tensor_slices(x_test,y_test)
db = db.shuffle(10000)

.map

• 資料預處理

def preprocess(x,y):
    x = tf.cast(x,dtype=tf.float32)/255.
    y = tf.cast(y,dtype=tf.int32)
    y = tf.one_hot(y,depty=10)
    return x,y

db2 = db.map(preprocess)

res = next(iter(db2))
res[0].shape,res[1].shape

res[1][:2]

.batch

• 一次性得到多張照片

db3 = db2.batch(32)
res = next(iter(db3))
res[0].shape,res[1].shape

db_iter = iter(db3)
while True:
    next(db_iter)

.repeat()

# 迭代不退出
db4 = db3.repeat()
# 迭代兩次退出
db3 = db3.repeat(2)

For example

def prepare_mnist_features_and_labels(x,y):
    x = tf.cast(x,tf.float32)/255.
    y = tf.cast(y,tf.int64)
    return x,y

def mnist_dataset():
    (x,y),(x_val,y_val)=datasets.fashion_mnist.load_data()
    y  =tf.one_hot(y,depth=10)
    y_val = tf.one_hot(y_val,depth=10)
    
    ds = tf.data.Dataset.from_tensor_slices((x,y))
    ds = ds.map(prepare_mnist_features_and_labels)
    ds =  ds.shffle(60000).batch(100)
    ds_val = tf.data.Dataset.from_tensor_slices((x_val,y_val))
    ds_val = ds_val.map(prepare_mnist_features_and_labels)
    ds_val = ds_val.shuffle(10000).batch(100)
    return ds,ds_val