Hmisc
Hmisc copied to clipboard
harrelfe
do not export 'recode'
If recode is an internal, undocumented function in Hmisc as it appears to be, it shouldn't be exported, especially considering it overwrites the more widely used car::recode.
E.g.
http://stackoverflow.com/questions/9337716/how-do-i-use-functions-in-one-r-package-masked-by-another-package
https://stat.ethz.ch/pipermail/r-help/2010-June/242466.html
recode is not an internal function, and is exported by Hmisc, 15 years before the existence of car. The bigger question is whether anyone is still using Hmisc's recode, i.e., whether it should stay in the package.
Since it's undocumented and bringing up its help dialog gives us "These are not to be called by the user or are undocumented." I doubt many casual users dare use it.
Good point; didn't remember I had never written a help file. Will remove from the package for the next version.
Thanks! This was my first R gotcha and it still trips me up sometimes :-)
On 05 Dec 2013, at 17:58, Frank Harrell [email protected] wrote:
Good point; didn't remember I had never written a help file. Will remove from the package for the next version.
— Reply to this email directly or view it on GitHub.
I do use it directly (usually with Hmisc:::recode after loading car second) and its disappearance bit my legacy code today. Found it in the prior code (https://github.com/harrelfe/Hmisc/blob/3cf7fce26cc1bf9728fdfc67164b5e24bf6bb7bc/R/recode.s) - agree that removing it (or renaming it) is probably best because it clashes with car so often even though you came first!
The removal of this function was announced earlier in the NEWS file.
You can find it in older versions of the package or I can send it to you
if needed.
Frank
On 09/09/2014 08:46 AM, Beau Bruce wrote:
I do use it directly (usually with Hmisc:::recode after loading car second) and its disappearance bit my legacy code today. Will find it in the prior code...
— Reply to this email directly or view it on GitHub https://github.com/harrelfe/Hmisc/issues/1#issuecomment-54970821.
Frank E Harrell Jr Professor and Chairman School of Medicine
Department of *Biostatistics* *Vanderbilt University*
Thanks - I found it - just commented since the issue was still open (even though I gather it is closed - in a fashion I agree with). Thanks for your help - I love Hmisc and rms.