Add missing shape parameter to numpy methods

[mbyrnepr2]

Add the missing shape parameter to the brain for:

• numpy.zeros_like
• numpy.full_like
• numpy.ones_like

Steps

• For new features or bug fixes, add a ChangeLog entry describing what your PR does.
• Write a good description on what the PR does.

Description

Type of Changes

	Type
✓	🐛 Bug fix
	✨ New feature
	🔨 Refactoring
	📜 Docs

Related Issue

Closes with relevant test pylint-dev/pylint#5871

[DanielNoord]

@mbyrnepr2 👍
Could you add a test for this?

[mbyrnepr2]

@DanielNoord sure thing. I am less familiar with astroid code so I will take a look at the test structure for this hopefully later today. If you can give me any pointers I would appreciate.

[DanielNoord]

@DanielNoord sure thing. I am less familiar with astroid code so I will take a look at the test structure for this hopefully later today. If you can give me any pointers I would appreciate.

I'm on mobile so I can't look up code easily right now.
Searching for numpy or np should probably show the other tests for this brain. Often we simply test that the thing we're inferring is correct.
Note: some tests use next(node.inferred()). We're trying to avoid this as this doesn't show whether there is ambiguity in the inference (which we often don't want). So it's often better to cast inferred() to a list and check that the length of that list is 1.

[mbyrnepr2]

@DanielNoord sure thing. I am less familiar with astroid code so I will take a look at the test structure for this hopefully later today. If you can give me any pointers I would appreciate.

I'm on mobile so I can't look up code easily right now. Searching for numpy or np should probably show the other tests for this brain. Often we simply test that the thing we're inferring is correct. Note: some tests use next(node.inferred()). We're trying to avoid this as this doesn't show whether there is ambiguity in the inference (which we often don't want). So it's often better to cast inferred() to a list and check that the length of that list is 1.

Great that's good to know, will try it out later @DanielNoord

[Pierre-Sassoulas]

Thank you for your First astroid contribution @mbyrnepr2 ;)

[cdce8p]

I am less familiar with astroid code so I will take a look at the test structure for this hopefully later today. If you can give me any pointers I would appreciate.

Adding another item here should work.
https://github.com/PyCQA/astroid/blob/29dcc35ac14ac21fe97502ac890a1faac936054e/tests/unittest_brain_numpy_core_numeric.py#L28-L29

[mbyrnepr2]

I am less familiar with astroid code so I will take a look at the test structure for this hopefully later today. If you can give me any pointers I would appreciate.

Adding another item here should work.

https://github.com/PyCQA/astroid/blob/29dcc35ac14ac21fe97502ac890a1faac936054e/tests/unittest_brain_numpy_core_numeric.py#L28-L29

Thanks @cdce8p! I was looking at this one earlier. It seems that the tests will pass with or without adding the parameter to this line; but that seems to be the structure for much of the numpy tests that I've seen.
So, for example, I have it set to ("zeros_like", "[1, 2]", "None", "K", "True", "None"), in my env. but the results are the same if I leave these optional parameters out also.

[mbyrnepr2]

Thank you for your First astroid contribution @mbyrnepr2 ;)

Thanks for the warm welcome, nice to be here!

[DanielNoord]

Change itself looks good to me!

[cdce8p]

Added some last changes in aeb1bf8. Looks good now!

Although not strictly necessary as astroid can't infer numpy natively anyway, I added the >1.17.0 requirement. According to the docs, that's where shape was added.
https://numpy.org/doc/stable/reference/generated/numpy.zeros_like.html

[mbyrnepr2]

Added some last changes in aeb1bf8. Looks good now!

Although not strictly necessary as astroid can't infer numpy natively anyway, I added the >1.17.0 requirement. According to the docs, that's where shape was added. https://numpy.org/doc/stable/reference/generated/numpy.zeros_like.html

Thank you for this @cdce8p!

[cdce8p]

And it's merged. Congrats on your first astroid contribution @mbyrnepr2 🚀