Specify stacklevel for warnings.warn for more helpful warning message.

[harupy]

Motivation

By default (stacklevel=1), a warning raised by warnings.warn generates a message like below.

> python test.py
/path/to/optuna/_experimental.py:86: ExperimentalWarning:

HyperbandPruner is experimental (supported from v1.1.0). The interface can change in the future.

This message shows the line of code that the warning originates from rather than where the deprecated function or class is called.

This PR fixes it by specifying stacklevel for warnings.warn. After the fix, a warning message should look like:

/path/to/file_that_calls_HyperbandPruner.py:6: ExperimentalWarning:

HyperbandPruner is experimental (supported from v1.1.0). The interface can change in the future.

https://docs.python.org/2/library/warnings.html#warnings.warn

Description of the changes

Specify stacklevel in warnings.warn.

[codecov-commenter]

Codecov Report

Merging #1419 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1419   +/-   ##
=======================================
  Coverage   88.04%   88.04%           
=======================================
  Files          98       98           
  Lines        7397     7397           
=======================================
  Hits         6513     6513           
  Misses        884      884           

Impacted Files	Coverage Δ
optuna/_deprecated.py	100.00% <ø> (ø)
optuna/_experimental.py	100.00% <ø> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4d6e1c7...cc45cf2. Read the comment docs.

[HideakiImamura]

Thanks for your contribution as always! It is reasonable to increase the stacklevel for decorators. LGTM.

[toshihikoyanase]

LGTM. Thank you for your contribution.

I confirmed that this change worked as expected using LightGBMTuner.best_booster, optuna.structs.TrialPruned, and GridSampler.