Hideaki Imamura

Document how each sampler treat pruned trials and failed trials

6

Motivated by useful discussions at https://github.com/optuna/optuna/issues/1552. # Motivation How each sampler in Optuna treat pruned trials and failed trials is different depending on each sampler. It is kind for users...

Add Other Performance Metrics

1

Hi! I would like to use the following performance metrics to validate the hyperparameter optimization algorithms with `kurobako`. - The number of budgets to reach the X% point of the...

[Optuna v3.0](https://github.com/optuna/optuna/releases/tag/v3.0.0) has been released, and it deprecated the old distributions and introduced the new distributions. We should replace the old definitions with new ones. ## What we should change....

Use `__future__.annotations` everywhere in the Optuna code base

21

### Motivation Optuna drops Python 3.6 from v3.1, so we can use `__future__.annotations`, which simplifies the code base. See [PEP 563](https://peps.python.org/pep-0563/), [PEP584](https://peps.python.org/pep-0584/), [PEP 585](https://peps.python.org/pep-0585/), and [PEP 604](https://peps.python.org/pep-0604/) for more details....

Speed up `_hypervolume.WFG`

2

### What is an issue? Optuna has implemented the _hypervolume.WFG class for calculating hypervolume, but this implementation is naive and slow. Speeding this up would be of great benefit in...

Document how each pruner handles `nan`

3

### What is an issue? Optuna has various pruners, and their handling of nan in intermediate values is different from each other. Checking the source code and documenting each handling...

### Motivation We are developing `optuna.terminator` as an automatic termination module, and have prepared `optuna.visualization.plot_terminator_improvement` for its visualization. However, there are inconsistencies in their behavior, and I propose that we...

Fix the documentation warnings with Sphinx `nitpicky`: `torch.distributed.ProcessGroup` is not linked to the pytorch document

2

### What is an issue? Derivative of https://github.com/optuna/optuna/issues/4213. Building the documentation using [Sphinx's nitpicky](https://www.sphinx-doc.org/en/master/usage/configuration.html#confval-nitpicky), we can observe several warnings. In this issue, I will suggest the problem reported [here](https://github.com/optuna/optuna/pull/4290#discussion_r1095256156). ####...

Fix the documentation warnings with Sphinx `nitpicky`: `pd.Dataframe` is not linked to the pandas document

1

### What is an issue? ### What is an issue? Derivative of #4213. Building the documentation using [Sphinx's nitpicky](https://www.sphinx-doc.org/en/master/usage/configuration.html#confval-nitpicky), we can observe several warnings. In this issue, I will suggest...

### What is an issue? Derivative of #4213. Building the documentation using [Sphinx's nitpicky](https://www.sphinx-doc.org/en/master/usage/configuration.html#confval-nitpicky), we can observe several warnings. In this issue, I will suggest to fix the problem of...