I have a 2d numpy array, matrix_a, shaped 2x20. How do I select columns 5:8 and columns 15:18.
matrix_a = np.array([[1,2,3,4,5...,19,20],
[1,2,3,4,5...,19,20]])
I want to select:
[[5,6,7,15,16,17],
[5,6,7,15,16,17]]
I can select the columns separately using matrix_a[0,5:8], is there a way to select it all at once?
Barring 0-index vs 1-index, in MATLAB, all I would have to do is:
matrix_a(:,[5:8, 15:18])
Is there an analogous, simple command in python?
You can use numpy.r_ to combine the two slices, which is a convenient way to Translate slice objects to concatenation along the first axis. And it can accept mixed lists, arrays, scalars and slice notation, e.g. np.r_[5, np.array([1,2]), [3]*2, 15:18] is a validation construction as well, which gives array([ 5, 1, 2, 3, 3, 15, 16, 17]):
matrix_a = np.tile(np.arange(1, 21), 2).reshape(2, 20)
matrix_a[:, np.r_[5:8, 15:18]]
#array([[ 6, 7, 8, 16, 17, 18],
# [ 6, 7, 8, 16, 17, 18]])
