[R-package] allow access to params in Booster

[jameslamb]

I've been trying to check LightGBM questions on Stack Overflow more frequently. A search today led me to this question from two years ago: https://stackoverflow.com/questions/49674112/accessing-lightgbm-model-parameters/50115384#50115384.

It pointed out that there's no way to access the parameters from Booster object. This is something that the Python package has had for a while.

Storing the parameters in the Booster can be helpful for cases where you've saved a model and then later load it in a session that doesn't have the original training code available. For example, you might want to be sure that the parameters you use for training continuation are the same as those you used in the original training. Or you might want to use parameters as metadata in a model monitoring system.

This pull request adds a new public attribute, Booster$params. Just like Booster.params in the Python package, it includes all passed-in parameters AND any parameters on the Dataset used for training (such as max_bin).

This PR also adds the first set of tests on saveRDS.lgb.booster() and readRDS.lgb.booster(), which were previously not covered by any tests.

[StrikerRUS]

(Shouldn't prevent to merge this PR) but I just want to say that a more general fix will be loading params from cpp side: #2613 (comment)

[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.