step() masked from base, but nonfunctional

[ichneumonid]

the step() function for stepwise deletion is masked, but it is not actually functional, at least in the glmmADMB package I had (and reloaded9/15/17 or so, to make sure it was up to date)

[bbolker]

a brief response: the problem is that step() is not a generic function (so it's hard to modify how it behaves for a particular model type), but was misbehaving in some bad way (TBH I no longer remember how/why) for glmmADMB fits.

Test out the fix in 332c179 (i.e. devtools::install_github("bbolker/glmmadmb") and reload) and see if it seems like an adequate solution?

[ichneumonid]

step() wasn't working for glmmADMB fits, it said it wasn't able to (I don't recall exactly what it said), but it also masked the same function from the base, so I couldn't use it elsewhere in the code where it would ordinarily work. I can't actually check that this fixed it, as I can't get the package to reload, despite using the same code I used to load it initially: install.packages("glmmADMB", repos=c("http://glmmadmb.r-forge.r-project.org/repos", getOption("repos")), type="source") library(glmmADMB) It gives me errors now: ERROR: compilation failed for package 'plyr' * removing 'C:/Users/lopre/OneDrive/Documents/R/win-library/3.4/plyr' ERROR: dependency 'stringi' is not available for package 'stringr' * removing 'C:/Users/lopre/OneDrive/Documents/R/win-library/3.4/stringr' ERROR: dependencies 'stringr', 'plyr' are not available for package 'glmmADMB' * removing 'C:/Users/lopre/OneDrive/Documents/R/win-library/3.4/glmmADMB' (And a whole bunch of warnings because of the non-loaded subpackages). So, in short, despite trying repeatedly, I can't tell if this fixed anything.

…

-Eric

On Mon, Sep 25, 2017 at 11:33 AM, Ben Bolker ***@***.***> wrote: a *brief* response: the problem is that step() is not a generic function (so it's hard to modify how it behaves for a particular model type), but was misbehaving in some bad way (TBH I no longer remember how/why) for glmmADMB fits. Test out the fix in 332c179 <332c179> and see if it seems like an adequate solution? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#8 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AeqAIDIlM7Lk0ub8ud1-IqwRagMb1mYMks5sl_HwgaJpZM4PedF-> .