Avraham Adler

No, I gave up a while ago. I should try again one of these days with @jeroen latest and greatest Rtools4. Sorry.

Hello. @jameslamb I get the following error: ```shell D:\LightGBM>git checkout fix/r-gpu hint: If you meant to check out a remote tracking branch on, e.g. 'origin', hint: you can do so...

True. None of those errors, but it is still finding the wrong Boost. ```shell D:\LightGBM>Rscript build_r.R --use-gpu --boost-root=D:/local/boost_1_80_0/ * checking for file 'D:/LightGBM/lightgbm_r/DESCRIPTION' ... OK * preparing 'lightgbm': * checking...

You may find the last line here of interest: ```shell D:\LightGBM>Rscript build_r.R --use-gpu --boost-librarydir=D:/local/boost_1_80_0/lib64-msvc-14.3/ * checking for file 'D:/LightGBM/lightgbm_r/DESCRIPTION' ... OK * preparing 'lightgbm': * checking DESCRIPTION meta-information ... OK...

Not to my knowledge.

Why don't you just find separate minimax interpolating polynomials for each interval?

I’m not in front of my computer, but if I recall correctly, some of the helper wrapper functions use the ellipse to pass parameters to the underlying `nloptr` call, so...

Since we started the process of switching the default order of the inequality, the code for all the `hin` and `heq` related functions have changed and so this PR should...

It has been over six years **and** we know there a suggestion that `nlopt` itself change some of the code. I don't think there is anything we can do in...

As mentioned above, this is something which would have to be exposed in NLOPT before anything can be done in nloptr, so this probably a **wontfix** since we cannot fix...