Add shekel funcfix

[ZohaibHassan16]

Description

This PR adds the Shekel function to the library's collection of standard algebraic test functions.

Related Issues

issue #10

Type of Change

• [ x] [ENH] - New feature (non-breaking change adding functionality))

How was this solved?

Implementation: Created shekel_function.py inside src/surfaces/test_functions/algebraic/standard/test_functions_nd/.

Registration: Updated the following init.py files to ensure the function is properly exported and accessible via the library's API:

src/surfaces/test_functions/algebraic/init.py

src/surfaces/test_functions/algebraic/standard/init.py

src/surfaces/test_functions/algebraic/standard/test_functions_nd/init.py

Checklist

Required

• [x ] PR title includes appropriate tag: [BUG], [ENH], [DOC] or [MNT]
• [x ] Code passes make check (lint, format, isort)

Tests

• [x ] Tests added/updated for changes

Documentation

• [ x] Documentation added/updated (docstrings, user guide, examples)

Note: New features ([ENH]) typically require both tests and documentation.
Bug fixes ([BUG]) should include a regression test when possible.

Testing

It can be verified by instantiating it through the standard algebraic interface:

from surfaces.test_functions.algebraic import ShekelFunction

func = ShekelFunction()
print(func.search_space)

Additional Notes

This is a fresh Pull Request intended to replace previous one that was "accidentally" combined with the other one

[SimonBlanke]

It would fit better to create a new module "test_functions_4d", because this test-function does not have a n_dim parameter. There is also some code-quality test, that fails.

[ZohaibHassan16]

Done and cleaned the code @SimonBlanke

[SimonBlanke]

There are some tests failing. Could you explain why?

[ZohaibHassan16]

My bad @SimonBlanke , it's because ShekelFunction was simultaneously still sitting in the standard_functions_nd list in constructor. Should be all green now

[SimonBlanke]

You need to change this and the search-space. This part still looks like an n-d test function

[ZohaibHassan16]

Hey @SimonBlanke ,
I've updated ShekelFunction with the following changes:

• Explicit Unpacking: Replaced the loop in the evaluator with explicit x0 through x3 handling for better clarity.

• Search Space: Updated the _search_space method to explicitly define the four dimensions.

Just a lil note, I think it will fail on CI run because in test_spec_immutable_across_calls currently assumes all multi-dimenstional functions are 2D? Please correct me on that

[SimonBlanke]

Some tests are still failing. Please also do a local test run before pushing the new commits.

[ZohaibHassan16]

Tests are failing because test_spec_immutable_across_calls test currently uses a hardcoded fallback that assumes any multi-dimensional function is strictly 2D. It passes {"x0": 0.5, "x1": 0.5}. Since ShekelFunction is our first 4D function, it crashes with a KeyError: 'x2' when the evaluator looks for the remaining dimensions.

We can fix this by making it dim-agnostic.

Specifically:

# Old code: 
# _ = func({"x0": 0.5} if len(func.search_space) == 1 else {"x0": 0.5, "x1": 0.5})

# New code:
params = {key: 0.5 for key in func.search_space.keys()}
_ = func(params)

[SimonBlanke]

Tests are failing because test_spec_immutable_across_calls test currently uses a hardcoded fallback that assumes any multi-dimensional function is strictly 2D. It passes {"x0": 0.5, "x1": 0.5}. Since ShekelFunction is our first 4D function, it crashes with a KeyError: 'x2' when the evaluator looks for the remaining dimensions.

We can fix this by making it dim-agnostic.

Specifically:

# Old code: 
# _ = func({"x0": 0.5} if len(func.search_space) == 1 else {"x0": 0.5, "x1": 0.5})

# New code:
params = {key: 0.5 for key in func.search_space.keys()}
_ = func(params)

That is right! Currently we do not even have 3D test functions. Only 1D, 2D and ND. And the test is set on this test-function collection, which is not good. I think your change makes sense. Please implement it :-)

[ZohaibHassan16]

Done @SimonBlanke , spec consistency tests are now passing.

[SimonBlanke]

I would like to merge this soon, but there are changes in the .gitignore that create merge conflicts. There should not be any changes in the .gitignore in this PR.

[ZohaibHassan16]

Resolved @SimonBlanke

[SimonBlanke]

There are some test fails. It is an easy fix. Just bump the number.

[SimonBlanke]

Well done.

[ZohaibHassan16]

Thanks for your patience !!! These were my first two PRs ever on GitHub, so I learned a lot, especially about how not to be a dumbass in a large codebase.