NumPy Fancy Array Indexing

From NovaOrdis Knowledge Base
Jump to navigation Jump to search

Internal

Overview

Fancy indexing is a term adopted by NumPy to describe indexing using integer arrays.

In case of a bi-dimensional array, passing an unidimensional list containing row indices returns a bi-dimensional array where only the rows whose indices were provided are present, in the order in which the indices were specified in the selection array.

a =  np.array([[10, 10], [20, 20], [30, 30], [40, 40]])

array([[10, 10],
       [20, 20],
       [30, 30],
       [40, 40]])

a[[3, 0]]

array([[40, 40],
       [10, 10]])

Using negative indices selects rows from the end.