Review: AI Gen Examples for numpy.linalg.svdvals function

[ogidig5]

numpy/numpy@main...ogidig5:numpy:add-svdvals-example

Here's what reflects when the docs build:

Checks

spin build && python -m pip install . && spin docs && python tools/refguide_check.py --doctests

• All tests have passed
• A visual inspection of the generated docs is correct

Acceptance Criteria:

Close when the following are complete, in order:

• Reviewed by @bmwoodruff
• Any suggestions to fork have been merged.
• Reviewed by Numpy Mainter:

When all tasks are complete:

• Submitted as PR to NumPy.

[ogidig5]

Let me check the syntax of this function and make sure it runs well on the Jupyter Notebook. It should be a minor change.

[bmwoodruff]

Viewing the docs is not the same as running the doctests with python toos/refguide_check --doctests.

[ogidig5]

I really need to quickly meet up with you. Seems I'm missing something crucial in the process. I want to work on that through the day

[ogidig5]

I've been able to figure out the authentication issue. Here's the current changes pushed to my fork: numpy/numpy@main...ogidig5:numpy:add-svdvals-example

[bmwoodruff]

You'll want to update the images to be current. Just edit your original message, so the new image shows up front.
Just about everything looks good now, with one issue:

• I ran the doctests, and they failed.

Did you run the tests? In particular, did you run python tools/refguide_check.py --doctests. You must run all the tests. This should have failed on your machine as we didn't set a seed.

The tests failed because the matrix values are random. You have to add # may vary or # random after the output , otherwise the tests will fail. This is what the relevant code could look like.

...
    array([2.05175379, 0.75852089, 0.10363495]) # may vary
...
    array([[1.59357377, 0.2369699 , 0.03501063], # may vary
           [2.00939638, 0.39010965, 0.06555355]])
...
    array([2.35955923, 0.67914258, 0.15643921]) # may vary

Please run the doctests, and verify that the test fails before you make the change. If all you run is python tools/refguide_check.py --doctests, that isn't enough. You first have to build the docs, then install numpy, and then you can run that function. The entire string of commands you need is

spin build && python -m pip install . && spin docs && python tools/refguide_check.py --doctests && spin lint

AFTER you have verified the doctests failed, then make the change by adding # may vary. Then test them again (build numpy, install it, run the doc tests)

spin build && python -m pip install . && spin docs && python tools/refguide_check.py --doctests && spin lint

Then try changing # may vary to # random, and verify you pass the tests as well.

Glad you submitted this one. It's another thing I need to add to automation. I'll have to add a checker to see if "random" values are entered, and then add this to the first line of output.

[bmwoodruff]

Here is the error message I got:

[bmwoodruff]

I'm going to close this one. It isn't passing the tests, and the internship has ended. I think it's just missing the # may vary tag before it will pass the tests.