Numpy库基础___二

ndarray一个强大的N维数组对象Array

•ndarray的变换

• x.reshape(shape)重塑数组的shape，要求元素的个数一致，不改变原数组

x = np.ones((2,3,4),dtype=np.int32)
print(x)
#[[[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]
#
# [[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]]
#[[[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]
#
# [[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]]

a = x.reshape((3,8))
print(a)
#[[1 1 1 1 1 1 1 1]
# [1 1 1 1 1 1 1 1]
# [1 1 1 1 1 1 1 1]]

• x .resize(shape)和x.reshape(shape)功能一样，但修改原数组

　　　　只有resize会改变原数组！

x = np.ones((2,3,4),dtype=np.int32)
print(x)
#[[[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]
# [[1 1 1 1]
#  [1 1 1 1]
#  [1 1 1 1]]]
a = x.resize((3,8))
print(x)
#[[1 1 1 1 1 1 1 1]
# [1 1 1 1 1 1 1 1]
# [1 1 1 1 1 1 1 1]]
print(a)
#None

• x.swapaxes(a,b)将数组中n个维度中的两个维度调换

x = np.ones((2,3,4),dtype=np.int32)
print(x)
# [[[1 1 1 1]
#   [1 1 1 1]
#   [1 1 1 1]]
# 
#  [[1 1 1 1]
#   [1 1 1 1]
#   [1 1 1 1]]]
x = x.swapaxes(0,2)
print(x)
#[[[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]]

• x.transpose(shape)矩阵x按照shape转置

x = np.ones((2,3,4),dtype=np.int32)
print(x)
#[[[1 1 1 1]
#   [1 1 1 1]
#   [1 1 1 1]]
# 
#  [[1 1 1 1]
#   [1 1 1 1]
#   [1 1 1 1]]]
x = x.transpose((2,1,0))
print(x)
[[[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]
# 
#  [[1 1]
#   [1 1]
#   [1 1]]]