[R-package] fix R examples and lgb.plot.interpretation()

[jameslamb]

*NOTE: Moving #2999 over here (based on a LightGBM branch) so we can test readthedocs builds. See #2999 (comment)

---

See comments from @StrikerRUS

• [R-package] [docs] Simplified examples to cut example run time (fixes #2988) #2989 (comment)
• [R-package] [docs] Simplified examples to cut example run time (fixes #2988) #2989 (comment)

A few of the R examples are broken right now. These weren't caught by tests because one is only a warning, one is a plot that is generated successfully but is incorrect, and one is an issue specific to our readthedocs builds.

I think the issue for lgb.cv() is #2715. In that issue we found that data.table 1.11.4 and R 3.6.0 lead to a data.table error ending in column or argument 2 is NULL. I think this can be fixed by upgrading to the data.table 1.12.1 in conf.py.

To fix the lgb.set.categorical() example, I changed file names across several examples to be sure that no two examples write to the same file name.

This issue with lgb.plot.interpretation() is because I made a change and forgot to use abs().

new plot:

I was not able to reproduce the data.table error in #2989 (comment). It

[StrikerRUS]

Building this branch right now.

[StrikerRUS]

It is failing due to wrong R version of r-data.table.

1.11.4 is the last version compiled for R 3.5.

[StrikerRUS]

• lgb.Dataset.set.categorical - FIXED: https://lightgbm.readthedocs.io/en/jameslamb-fix-r-examples/R/reference/lgb.Dataset.set.categorical.html
• lgb.plot.interpretation - FIXED: https://lightgbm.readthedocs.io/en/jameslamb-fix-r-examples/R/reference/lgb.plot.interpretation.html
• lgb.cv - FIXED: https://lightgbm.readthedocs.io/en/jameslamb-fix-r-examples/R/reference/lgb.cv.html

[StrikerRUS]

I believe we should bump version to 1.12 here

LightGBM/R-package/DESCRIPTION

Line 36 in 995a597

data.table (>= 1.9.6),

and correct the FAQ entry, because not only lgb.dl is affected
https://github.com/microsoft/LightGBM/blob/master/docs/FAQ.rst#3-error-in-datatabledatatableargument-2-is-null

[jameslamb]

I believe we should bump version to 1.12 here

LightGBM/R-package/DESCRIPTION

Line 36 in 995a597

data.table (>= 1.9.6),

and correct the FAQ entry, because not only lgb.dl is affected
https://github.com/microsoft/LightGBM/blob/master/docs/FAQ.rst#3-error-in-datatabledatatableargument-2-is-null

I disagree about updating the version floor. It seems that this is not specific to lgb.dl(), but I showed in #2715 (comment) that the incompatibility is more complicated than just "LightGBM doesn't work with 1.11.4". The 1.12 series is only about 15 months old, so I'm not prepared to force people to upgrade to it yet.

I do agree on updating the FAQ entry...I'll do that directly in this PR.

[jameslamb]

@StrikerRUS are you ok with the update I made to the FAQ? I think it came after your last approval.

[StrikerRUS]

@jameslamb I'm OK with that you are against bumping the required version. But the FAQ entry should state that in case of that error version upgrade is required anyway for someone who encoutrered it, I believe.

Sure, please do it right in this PR.

[jameslamb]

@jameslamb I'm OK with that you are against bumping the required version. But the FAQ entry should state that in case of that error version upgrade is required anyway for someone who encoutrered it, I believe.

Sure, please do it right in this PR.

ha I forgot to push the commit!! I pushed to origin (my fork) a few days ago...forgot this PR is on a LightGBM branch. My ask "are you ok with the update" probably was confusing, sorry!

Here it is: 0eb0243

[StrikerRUS]

Ah, I see! FAQ update looks good. Thanks a lot for fixing all issues!

[jameslamb]

Since this one is a user-facing fix (fixing lgb.plot.interpretation()) and a non-controversial one, I feel ok merging it without an approval from an R maintainer.

Thanks for the help @StrikerRUS !