Skip to content

Removed test failing only on specific devices only#862

Merged
jf-som merged 3 commits intodevelopfrom
fix/frequency_weighting_tests
Nov 14, 2025
Merged

Removed test failing only on specific devices only#862
jf-som merged 3 commits intodevelopfrom
fix/frequency_weighting_tests

Conversation

@jf-som
Copy link
Copy Markdown
Contributor

@jf-som jf-som commented Nov 12, 2025

Which issue(s) are closed by this pull request?

None

Changes proposed in this pull request:

As @f-brinkmann noticed, the function test_frequency_weighting_filter_errwgt_recommended in test_frequency_weighting.py fails on his MacOS device, but not on my Windows device or on circleCI.
Apparently, the scipy least-squares optimization can be platform/hardware-specific. I removed the failing assert, since it was only for a recommended function argument, and adjusted the recommendation in the docstring comment accordingly.

Moved to a new issue:

Do we have any mechanism to run tests on multiple operating systems and architectures? If not, maybe we should add a tool that runs tests on multiple setups at least before new releases.

…ers perform better overall and adapt comments
@jf-som jf-som added this to the v0.7.4 milestone Nov 12, 2025
@jf-som jf-som self-assigned this Nov 12, 2025
@jf-som jf-som added bug For bugs not on the master branch filter labels Nov 12, 2025
@jf-som jf-som changed the title removed check that high frequency emphasized frequency weighting filt… Removed test failing only on specific devices only Nov 12, 2025
Comment on lines 126 to 127
near the nyquist frequency. Doing so sometimes leads to better results
for typical sampling rates, but much worse for very high rates.
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

'Sometimes' sounds a little pessimistic and vague. How about being more specific, e.g.:

Suggested change
near the nyquist frequency. Doing so sometimes leads to better results
for typical sampling rates, but much worse for very high rates.
near the nyquist frequency. Note that the parameters may depend on the sampling
rate and operating systems.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Perhaps "results" would be more fitting than "parameters". I gave it a new try.

@jf-som jf-som requested a review from f-brinkmann November 13, 2025 08:30
@mberz
Copy link
Copy Markdown
Member

mberz commented Nov 13, 2025

Please do not use PRs to discuss feature planning beyond the scope of the PR.
You can open a new issue for this.

@jf-som
Copy link
Copy Markdown
Contributor Author

jf-som commented Nov 13, 2025

Please do not use PRs to discuss feature planning beyond the scope of the PR. You can open a new issue for this.

If you mean the automated multi-OS testing, I was going to open a new issue but wanted to ask if this was already solved and I just didn't know ^^
I will open a new issue for it then.

Copy link
Copy Markdown
Member

@mberz mberz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Spotted a small typo. Approved ;)

Co-authored-by: Marco Berzborn <mberz@users.noreply.github.com>
@jf-som jf-som merged commit 1bb2640 into develop Nov 14, 2025
2 of 8 checks passed
@jf-som jf-som deleted the fix/frequency_weighting_tests branch November 14, 2025 07:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug For bugs not on the master branch filter

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants