Kenshin Abe comments

Results 27 comments of


                                            Kenshin Abe

Refactor `_tell.py`

Thanks for the update! Almost LGTM, but you dropped the cast to `float` of each element of `values` by [this commit](https://github.com/optuna/optuna/pull/3841/commits/48883f731d5704009b2a1f592dba307e48facfef)?

Skip tests if required libraries are not installed

Thanks for the PR. The change looks great overall, but I'm afraid that, in the current PR, we might miss some import errors when we fail to build the CI...

`constant_liar=True` does not work in multi-objective `TPESampler`

To support this feature, we might need to be careful with #3676 and #3677, which are inf-related issues with multi-objective `TPESampler`.

Unify optimization history tests

Thanks for the careful review! I updated several points (described in each commit message) including your comment. PTAL 🙏

Gather mask of None parameter in TPESampler

@HideakiImamura Although I haven't modified the test yet, could you briefly check if this strategy is okay when you have time? By this PR, `_sample_relative` and `sample_independent` will get a...

Gather mask of None parameter in TPESampler

Thanks! I fixed the error. Could you take a look again and assign another reviewer?

Refactor `test_pareto_front`

I reran the CI and now it's running fine! Let me also review the code.

Refactor `test_pareto_front`

Thanks! Almost LGTM, but could you reflect [this comment](https://github.com/optuna/optuna/pull/3798#discussion_r936386238)?

Add issue template for code fix

Thanks for the comment! That makes sense to me, but I made the name consistent with the existing label ["code-fix"](https://github.com/optuna/optuna/labels/code-fix) ( "Change that does not change the behavior, such as...

Add issue template for code fix

(Sorry, I just deleted a commit that I made by mistake. I didn't make any change by the above force-push.)