You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Implementation-wise, this may be challenging (w/o just telling users to devtools::install_github('GLEON/GLMr',ref='v3.1.2') or whatever.
Maybe a way to pull other versions down from an archive somewhere and pop them in. The package is already big, so we probably don't want to put any more versions in the package itself.
Relevant because of recent issues w/ @hdugan's model runs
The text was updated successfully, but these errors were encountered:
I've thought of this a little. What you could actually do if you wanted to be fancy is include the binaries in compressed form, and decompress them for running. The compressed set of different versions is only slightly larger than the individual model because of the similarity between the binaries. That would keep the size down.
It gets slightly annoying on Mac and Linux though as the execution permission has to be set on the binary which might get tricky to deal with on the fly (sys calls to chmod or something, ugh). But still, could be done. Uncompress binary into temp directory and run. Seems easy enough. Do we get into a weird situation with the NML file? Do we need to version those too?
"Do we get into a weird situation with the NML file? Do we need to version those too?"
yes, I would think
Maybe do something like: Include the most recent public release, and the previous minor release (so in this case, that would be v2.1.0 of GLM). Getting other releases would require the simple devtools::install_github..., which isn't that much of a pain, but just requires a little heavier dependencies.
Implementation-wise, this may be challenging (w/o just telling users to
devtools::install_github('GLEON/GLMr',ref='v3.1.2')
or whatever.Maybe a way to pull other versions down from an archive somewhere and pop them in. The package is already big, so we probably don't want to put any more versions in the package itself.
Relevant because of recent issues w/ @hdugan's model runs
The text was updated successfully, but these errors were encountered: