Select 使用来自 numpy 中的 ndarray 的索引的多个 arrays

Question

Say if I have a 2D array:

y = np.arange(35).reshape(5,7)
# array([[ 0,  1,  2,  3,  4,  5,  6],
#        [ 7,  8,  9, 10, 11, 12, 13],
#        [14, 15, 16, 17, 18, 19, 20],
#        [21, 22, 23, 24, 25, 26, 27],
#        [28, 29, 30, 31, 32, 33, 34]])

and select the 2nd and 3rd elements of the 1st, 3rd and 5th array like so:

y[np.array([0,2,4]), 1:3]
# array([[ 1,  2],
#        [15, 16],
#        [29, 30]])

I cannot find a way to replicate this using arrays in place of the slice for indexing, the following doesn't work, I must be able to use arrays to index as I sometimes might be interested in the 2nd and 4th elements of the arrays and so on:

y[np.array([0,2,4]), np.array([1,2])]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
IndexError: shape mismatch: indexing arrays could not be broadcast together with shapes (3,) (2,)

How can I achieve my desired functionality?

Answer 1

np.ix_() is designed for this type of problem.

def getSub():
    y = np.arange(35).reshape(5,7)
    locs = np.ix_([0,2,4],[1,2])
    return y[locs]


>>> getSub()
array([[ 1,  2],
       [15, 16],
       [29, 30]])

Answer 2

y[np.array([0,2,4]), np.array([[1],[2]])].T

Answer 3

you can try to use this y[np.array([[0,2,4]]*2),np.array([[1]*3,[2]*3])].T

Answer 4

Quick, dirty way to achieve it with double indexing.

y[np.array([0,2,4]),:][:,np.array([1,2])]