np.random.randn生成随机矩阵的规模特点

测试程序：

import numpy as np
A=np.random.randn(4,3)
B=np.random.randn(5,4,3)
print(A,A.shape)
print(B,B.shape)

测试结果：

[[ 1.65116631  0.65433859  0.60835888]
 [-0.5105315   1.27626974 -1.17579392]
 [-0.21566487 -0.57976917  1.59900943]
 [ 1.00468399 -0.38331758  1.49912217]] (4, 3)
[[[-0.54863125  1.26576023 -2.95168856]
  [ 0.27166768 -0.90973455 -1.70130229]
  [ 1.4344519   0.46945974 -0.51778667]
  [ 0.82859325 -2.55669375 -1.15995574]]

 [[-1.867543    1.26659434  1.09745486]
  [-0.68603883  0.83188864  0.01995746]
  [ 0.10028617  0.55817263 -1.10271119]
  [ 0.17845894  0.05393899 -0.74882348]]

 [[-0.54340926 -0.36803138  1.70142752]
  [-1.56940267 -0.45341488  0.25691709]
  [ 1.00348669 -0.30357184  0.11269915]
  [-0.75230232 -0.66815715  0.1445032 ]]

 [[-1.4450306   0.32532832 -0.84110991]
  [ 0.57615025  2.47953269  0.47464932]
  [-0.45915104  1.09317274 -1.26994611]
  [ 2.74141194  1.52149005 -0.08431544]]

 [[-0.57330565 -0.77312988 -0.35378862]
  [-0.74788065  1.34449482 -0.04547778]
  [-0.86599271 -0.48533614  0.08144477]
  [ 0.82996177  1.18632351  0.10067105]]] (5, 4, 3)

Process finished with exit code 0