[R-package] change default nrounds to 100 to match LightGBM core library default

[david-cortes]

Fixes #4196

The R interface has some default values for arguments related to training and evaluation which differ from Python, namely:

• Number of boosting rounds (100 in python, 10 in R).
• Number of cross-validation folds (5 in python, 3 in R).

This PR changes the default values in R to match those of Python.

[ghost]

All CLA requirements met.

[jameslamb]

Thanks very much for your interest in LightGBM and attention to detail! Please see my comment below about nfold.

[jameslamb]

Thanks for this!

The failing CI job is not related to your changes. I've manually restarted the build, and will merge this when it passes.

[jameslamb]

Before I merge, I want to check. @StrikerRUS are you ok with this change?

[StrikerRUS]

I'm totally fine with this fix! I remember I spent quite large amount of time syncing Python, native and docs param values several years ago. I'm glad that someone has started working of R-package now!
I'd like to note that the next release of LightGBM will be major 4.0.0, so we're free to change anything we need/want.

[jameslamb]

Ok thanks!

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.