[BUG] ensure BaseMetaObject resets at set_params; add 'test_set_params_resets_fitted_state' test to trigger error#467
Conversation
…Blanke/skbase into bugfix/reset_at_set_params
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #467 +/- ##
==========================================
+ Coverage 85.07% 86.28% +1.21%
==========================================
Files 45 50 +5
Lines 3015 3602 +587
==========================================
+ Hits 2565 3108 +543
- Misses 450 494 +44 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
The tests I added to trigger the bug described in the issue now pass after implementing my fix. |
BaseMetaObject resets at set_params; add 'test_set_params_resets_fitted_state' test to trigger error
There was a problem hiding this comment.
Brilliant!
I now understand that the return self call is the culprit.
This is unfortunate, I did spot it as an inconsistency and indeed it actually turned out (by your diagnosis) that it is the problem here.
If the clause if params is None: return self would not happen, would the issue not have happened? That is a strong indication that this special case caused a problem.
Either way, that is out of scope for this PR, but perhaps we should open an issue on whether this needs a change?
|
issue opened here: #469 |
Reference Issues/PRs
fixes #412
What does this implement/fix? Explain your changes.
This PR adds a test to trigger the bug described in #412. If desired this PR will also add a fix to the bug.
Does your contribution introduce a new dependency? If yes, which one?
What should a reviewer concentrate their feedback on?
Any other comments?
PR checklist
For all contributions
the PR topic is related to enhancement, CI/CD, maintenance, documentation, or a bug.
For code contributions