Thank you for reporting this, @ml-practitioner . I propose using the 3-by-2 array for all examples and fixing the code. Having a non-square array helps with the comprehension here, in my opinion.

For the 4th figure, the non-sequential values (I think) are also meant to create a non-trivial example. In that case, it might be better to have data = np.array([[1, 2], [5, 3], [4, 6]]) throughout these 4 examples, so we will need to update the figures. @MarsBarLee can help with that.

Let me know if this looks like a reasonable solution.

I agree with both of your suggestions. If you're going to update the figures, might I also suggest orienting the "2 5 6" on the right side of the 4th figure horizontally instead of vertically? The vertical orientation implies the result is [3, 1]-shaped rather than its actual [3,]-shape.

This can be a very confusing topic for beginners, so thank you for addressing this. Oh, and I see now that I incorrectly stated that the answer for the left-hand side of the 4th figure was [2, 5, 6]--the correct answer is indeed [5, 6] as shown.

I am interested in updating the figures for this issue and would appreciate some direction regarding the accepted process for creating these PNG images.

Hi @AlexanderPatino, thank you so much for your interest in working on this issue! However Melissa and I are already working on updating these graphics. You can check out our NumPy Documentation tracking issue here to see on other issues you can help out on. In terms of the process for creating images, we're looking to create SVGs which can be done on graphics software such as Inkscape. If you have any questions, please let us know in the tracking issue :)