Numpy array views syntax

Question

I think I understand the basics of numpy array views (it doesn't copy data and we're basically referring to the same data buffer as the array but with different offsets, etc.). I am however confused about the syntax.

What I do understand is that if I were to run this:

import numpy as np
a = np.random.random_sample((100,100))
b = a[:10, :10]

Then b would be the first 10 rows and the first 10 columns of a.

What I don't understand is what a[::10] does. Could someone please explain that?

Also - are there any other similar things I should know about when dealing with numpy arrays?

I've looked at array views on http://docs.scipy.org/doc/numpy/reference/generated/numpy.ndarray.view.html but it doesn't really help me.

Adam Hughes · Accepted Answer · 2015-01-24 23:45:42Z

2

The ::10 syntax means sample the entire range returning every ten elements. SO if you wanted to return 1/100th of the data (ie every 10th row, every 10th column), you'd do:

b = a[::10, ::10]

a[::10] is just shorthand for a[::10, ::] if I recall.

answered Jan 24, 2015 at 23:45

Adam Hughes

16.5k14 gold badges100 silver badges140 bronze badges

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

5 Comments

Sanjay Manohar Over a year ago

isn't that 1/100th of the data?

Abhinav Ramakrishnan Over a year ago

okay... so what does a[:2:10, :2:10] do? it returns something when I try it out on IDLE?

Adam Hughes Over a year ago

What is the shape of the resulting array? I don't have access to python right now to test it. And is it any different than a[2:10, 2:10]

Abhinav Ramakrishnan Over a year ago

I figured it out. Turns out the syntax is array[start:end:steps, start:end:steps,..] for as many dimensions as I have. So basically in the example I gave with the shape being (100,100), a[::10, ::10] == a[:100:10, :100:10]

Adam Hughes Over a year ago

Interesting. Admittedly, I usually get tripped up with indexing in numpy and only choose to remember a subset of use cases. I've never run into any trouble with using the :: syntax exclusively.

Collectives™ on Stack Overflow

Numpy array views syntax

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related