在数组索引中自动切换维度（python）

我想用Python自动化下面的操作。给定一个n维

np.数组

，例如

ar = np.array([[[ 8, 10],
                [ 9, 11],
                [10, 12]],
               [[ 9, 11],
                [10, 12],
                [11, 13]]])

dim_names = ['A','B','C']

带有

尺寸名称中每个尺寸的名称，例如

ar = np.array([[[ 8, 10],
                [ 9, 11],
                [10, 12]],
               [[ 9, 11],
                [10, 12],
                [11, 13]]])

dim_names = ['A','B','C']

选择任何给定尺寸的所有边缘，例如

ind  = dim_names.index('B')

for i in range(ar.shape[ind]):
    print(ar[:,i,:])

但是根据
ind
索引必须改变，例如对于
ind=0
我们需要
ar[i，：，：]
或者对于
ind=2
ar[：，：，i]
。此外，它应该适用于任意数量的维度数组。
可能有更好的方法解决您的问题，但对于问题，您可以在循环中动态构造索引，即指定索引和切片的所需维度

沿第一维度切片：

ind = 0
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim             # inintialize the index with the same length 
                                              # as the dimensions
    idx[ind] = i                              # modify the index of the desired dimension
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]
# [10 12]]
#[[ 9 11]
# [10 12]
# [11 13]]

ind = 1
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]]
#[[ 9 11]
# [10 12]]
#[[10 12]
# [11 13]]

ind = 2
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8  9 10]
# [ 9 10 11]]
#[[10 11 12]
# [11 12 13]]

沿第二维度切片：

ind = 0
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim             # inintialize the index with the same length 
                                              # as the dimensions
    idx[ind] = i                              # modify the index of the desired dimension
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]
# [10 12]]
#[[ 9 11]
# [10 12]
# [11 13]]

ind = 1
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]]
#[[ 9 11]
# [10 12]]
#[[10 12]
# [11 13]]

ind = 2
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8  9 10]
# [ 9 10 11]]
#[[10 11 12]
# [11 12 13]]

沿三维切片：

ind = 0
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim             # inintialize the index with the same length 
                                              # as the dimensions
    idx[ind] = i                              # modify the index of the desired dimension
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]
# [10 12]]
#[[ 9 11]
# [10 12]
# [11 13]]

ind = 1
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8 10]
# [ 9 11]]
#[[ 9 11]
# [10 12]]
#[[10 12]
# [11 13]]

ind = 2
​
for i in range(ar.shape[ind]):
    idx = [slice(None)] * ar.ndim
    idx[ind] = i
    print(ar[tuple(idx)])

#[[ 8  9 10]
# [ 9 10 11]]
#[[10 11 12]
# [11 12 13]]