如何在 numpy.where 条件后获取特定数组 position？

Question

I am using the where condition to my array and get back the array position array_string[0,0] with 'A' but I need the position array_string[0,1] with 200.

array_string = np.array([['A', 200],['B', 100]])

array_string[ np.where( array_string == 'A' ) ]

Answer 1

First of all, there is no need to use the np.where . Secondly, you can get your desired answer like this:

array_string[array_string[0,:]=='A', :]

> array([['A', '200']], dtype='<U3')

Translated into human language: check the first row in all columns, and if it is equal to A, get me the entire row (all the columns of this row).

Answer 2

This happens, because np.where return indices for each dimension where element was found.

array_string = np.array([['A', 200], ['B', 100]])

rows_found, cols_found = np.where(array_string == 'A')
rows = array_string[rows_found]  # returns [['A', 200]]
print(rows[0][1])  # prints 200


# Another example
array_string = np.array([['A', 200], ['B', 100], ['A', 500]])

rows_found, cols_found = np.where(array_string == 'A')
rows = array_string[rows_found]  # returns [['A' 200], ['A' 500]]
print(rows[:, 1])  # prints [200, 500]