Python/Numpy: Selecting a Specific Column in a 2D Array

I’ve been playing around with numpy this evening in an attempt to improve the performance of a Travelling Salesman Problem implementation and I wanted to get every value in a specific column of a 2D array.

The array looked something like this:

>> x = arange(20).reshape(4,5)
>>> x
array([[ 0,  1,  2,  3,  4],
       [ 5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14],
       [15, 16, 17, 18, 19]])

I wanted to get the values for the 2nd column of each row which would return an array containing 1, 6, 11 and 16.

For some reason I was expecting it to be quite complicated but in fact we can do this by using matrix style syntax like so:

>>> x[:, 1]
array([ 1,  6, 11, 16])

Here we are first saying that we want to return all the rows by specifying ‘:’ and then the ’1′ indicates that we only want to return the column with index 1.

If we wanted to return a specific row as well then we’d specify a value before the comma and it’d be a standard 2D array value lookup:

>> x[2,1]
11

or

>> x[2][1]
11