[docs] Clarify the fact that predict() on a file does not support saved Datasets (fixes #4034)

[jameslamb]

Fixes #4034.
Contributes to #4310.

predict() methods in LightGBM can be used with data stored in a text file (CSV, TSV, or LibSVM). Those methods do not support creating predictions on other formats like LightGBM Dataset objects saved in a binary file.

This PR proposes the following changes to fix #4034:

• clarifies the supported file formats for predict methods in the R and Python package documentation
• adds a unit test in the R package on the behavior of calling Booster$eval() on a constructed Dataset stored in a file ([R-package] predict() breaks when using a Dataset stored in a file #4034 (comment))
• updates the relevant error message in the C++ code used to determine file type when predicting on a file, to make it clearer that only certain formats of text file are supported

[StrikerRUS]

Thank you for picking this up! I have some comments below:

[StrikerRUS]

Thank you! LGTM, except one minor unifying suggestion.

[StrikerRUS]

Let's make identical descriptions really identical without paraphrasing. Users might spend their time trying to find a difference where there is actually no any difference. Also, it will help us to not miss all occurrences of identical parameters during possible future updates.

Suggested change

	#' a character representing a path to a text file (CSV, TSV, or LibSVM)
	#' or a LightGBM Dataset binary file
	#' a character representing a path to a text file (CSV, TSV, or LibSVM),
	#' or a character representing a path to a binary \code{Dataset} file

[jameslamb]

oh sure, seems fine to me!

I'll do this one manually instead of applying in the browser, since the corresponding lgb.Dataset.Rd will also have to be regenerated.

[jameslamb]

updated in de189a6, thanks as always for being thorough! You're right, that will make it easier to catch all such phrases in the future.

[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.