contramundum53

Of course, thank you!

Sorry, we are having some discussions about this issue and #5060 among the core develop members right now. The core dev members are concerned that putting _CachedStorage class inside RDBStorage...

It was my mistake to make this issue contribution-welcome. I'll add needs-discussion flag here. Sorry for anybody who have spent time working on this issue.

Thank you for your suggestion! How about [Parallel Coordinate Plot](https://optuna--5053.org.readthedocs.build/en/5053/reference/visualization/generated/optuna.visualization.plot_parallel_coordinate.html#optuna.visualization.plot_parallel_coordinate)? Do you think beeswarm plot looks better?

By the way, I like how it avoids putting many points in the same location. Do you think it is easy enough to do with plotly?

Sorry I think I'm still not fully understanding your intended plot. > parameter values on y axis and function value on x Do you mean that the colors represent the...

Thanks for your suggestion! For the specific examples given above, you can simply call `study.optimize` twice with different samplers, like ``` study = optuna.create_study(sampler=SAMPLER1) study.optimize(objective, n_trials=n_trials_for_first_stage) study.sampler = SAMPLER2 study.optimize(objective,...

Another situation is that you want to sample with SAMPLER1 and SAMPLER2 alternatively (or with a specified probability). In this case, I recommend you create two studies with a shared...

Thank you for your suggestion! You can change your plot to your favorite color scale by the following code: ``` fig = optuna.visualization.plot_parallel_coordinate(study) fig.data[0].line["colorscale"] = "RdYlBu" # Reverse color scale...

The story for the curent behavior of the color is a bit long: When we only had single-objective optimization, we first wanted to show good lines in dark colors and...