2D Numpy Array indexing by using 2D array [duplicate]

Question

  position_array = np.array([[1,2]
                      [0,1]
                      [0,2]])

  original_array = np.array([[5,6,7]
                             [7,8,2]
                             [10,6,2]])

result_i_want = np.array([[6,7]
                        [7,2]
                        [10,2]])

how to do this?

get a new array from original_array and use position_array as index ??

for example when the first line of position_array is [1,2] that mean get column 1 and column 2 from the first line of original_array

how to do this simplest way?

yatu · Accepted Answer · 2019-08-31 10:36:31Z

2

You want np.take_along_axis, setting axis=1:

np.take_along_axis(original_array, position_array, axis=1)

array([[ 6,  7],
       [ 7,  8],
       [10,  2]])

Or you can also use advanced indexing:

original_array[np.arange(original_array.shape[0])[:,None], position_array]

array([[ 6,  7],
       [ 7,  8],
       [10,  2]])

answered Aug 31, 2019 at 10:36

yatu

88.6k12 gold badges93 silver badges148 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

2D Numpy Array indexing by using 2D array [duplicate]

1 Answer 1

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Linked

Related