tensorflow学习 数据的拆分与拼接 tf.split() tf.concat() numpy.split() np.concatenate()

用法如以下代码

import tensorflow as tf
import numpy as np

data = np.array([
    [1,2,3,4,5,6,7,8,9,0],
    [11,12,13,14,15,16,17,18,19,1],
    [21,22,23,24,25,26,27,28,29,30],
    [31,32,33,34,35,36,37,38,39,40]
])

sess = tf.InteractiveSession()
tf.global_variables_initializer().run()

# 按行切分（按第一维度）
data1,data2 = tf.split(data,[2,2],0)

row1 =  data1.eval()
row1

array([[ 1,  2,  3,  4,  5,  6,  7,  8,  9,  0],
       [11, 12, 13, 14, 15, 16, 17, 18, 19,  1]])

row2 = data2.eval()
row2

array([[21, 22, 23, 24, 25, 26, 27, 28, 29, 30],
       [31, 32, 33, 34, 35, 36, 37, 38, 39, 40]])

# 按列切分（第二维度）
col1,col2 = tf.split(data,[5,5],1)

col1_eval = col1.eval()
col1_eval

array([[ 1,  2,  3,  4,  5],
       [11, 12, 13, 14, 15],
       [21, 22, 23, 24, 25],
       [31, 32, 33, 34, 35]])

col2_eval = col2.eval()
col2.eval()

array([[ 6,  7,  8,  9,  0],
       [16, 17, 18, 19,  1],
       [26, 27, 28, 29, 30],
       [36, 37, 38, 39, 40]])

# 在列方向上拼接
data1 = tf.concat([row1,row2],axis=1)
data1.eval()

array([[ 1,  2,  3,  4,  5,  6,  7,  8,  9,  0, 21, 22, 23, 24, 25, 26,
        27, 28, 29, 30],
       [11, 12, 13, 14, 15, 16, 17, 18, 19,  1, 31, 32, 33, 34, 35, 36,
        37, 38, 39, 40]])

# 在行方向上拼接
data2 = tf.concat([col1_eval,col2_eval],0)
data2.eval()

array([[ 1,  2,  3,  4,  5],
       [11, 12, 13, 14, 15],
       [21, 22, 23, 24, 25],
       [31, 32, 33, 34, 35],
       [ 6,  7,  8,  9,  0],
       [16, 17, 18, 19,  1],
       [26, 27, 28, 29, 30],
       [36, 37, 38, 39, 40]])

# numpy 版本  列方向上（第二维度）均分
data_np_sp = np.split(data,2,axis=1)
data_np_sp

[array([[ 1,  2,  3,  4,  5],
        [11, 12, 13, 14, 15],
        [21, 22, 23, 24, 25],
        [31, 32, 33, 34, 35]]), array([[ 6,  7,  8,  9,  0],
        [16, 17, 18, 19,  1],
        [26, 27, 28, 29, 30],
        [36, 37, 38, 39, 40]])]

#列方向（第二维度）可以不均分，列表的第一个数量根据情况变化
data_np_asp = np.array_split(data,3,axis=1)
data_np_asp

[array([[ 1,  2,  3,  4],
        [11, 12, 13, 14],
        [21, 22, 23, 24],
        [31, 32, 33, 34]]), array([[ 5,  6,  7],
        [15, 16, 17],
        [25, 26, 27],
        [35, 36, 37]]), array([[ 8,  9,  0],
        [18, 19,  1],
        [28, 29, 30],
        [38, 39, 40]])]

# 在行方向上拼接（第一维）
data_np_con_0 = np.concatenate(np.array(data_np_sp),axis = 0)
data_np_con_0

array([[ 1,  2,  3,  4,  5],
       [11, 12, 13, 14, 15],
       [21, 22, 23, 24, 25],
       [31, 32, 33, 34, 35],
       [ 6,  7,  8,  9,  0],
       [16, 17, 18, 19,  1],
       [26, 27, 28, 29, 30],
       [36, 37, 38, 39, 40]])

# 在列方向上拼接
data_np_con_1 = np.concatenate((data_np_asp[0],data_np_asp[1],data_np_asp[2]),axis = 1)
data_np_con_1

array([[ 1,  2,  3,  4,  5,  6,  7,  8,  9,  0],
       [11, 12, 13, 14, 15, 16, 17, 18, 19,  1],
       [21, 22, 23, 24, 25, 26, 27, 28, 29, 30],
       [31, 32, 33, 34, 35, 36, 37, 38, 39, 40]])