====== GNU R ======


The [[https://www.r-project.org/|GNU R programming language]] is often paired with [[https://posit.co/download/rstudio-desktop/|posit's RStudio Desktop IDE]] and a suit of libraries (language packages) known as [[:tools:tidyverse|tidyverse]] (includes ''dplyr'' and ''ggplot2''). 

You can also find a web interface to [[http://rhea.wpic.upmc.edu:8787/|Rstudio on rhea]] (also [[:admin:remoteaccess]]).

[[:tools:tutorials|Tutorials]] contains additional resources.

See [[:tools:r:issues]] for log of debugged problems.
===== Notes =====
==== na.action ====
When adding module residuals back to a dataframe, you need ''residuals()'' to return the same length as the input data.frame. use ''lm(na.action=na.exclude)''.

For example,

<code=R>
d <-  data.frame(x=c(1:4,NA),y=1:5);
m <- lm(x~y,d,na.action=na.exclude);
nrow(d);              # 5
length(m$residuals);  # 4
length(residuals(m))  # 5
</code>


This avoids the error

  Error in `$<-.data.frame`(`*tmp*`, ...
    replacement has 237 rows, data has 348


==== MASS::select vs dplyr::select  ====
If you load MASS after dplyr, ''select'' will be ''MASS::select'' not ''dplyr::select'' and you're likely to encounter hard-to-debug errors about unused arguments

> Error in select ...  : unused arguments
 
**solutions** include 
  - load MASS first
  - force select to be ''dplyr'''s version, or 
  - unload MASS if you don't need it

<code>
# load mass before dplyr to have 'select' be from dplyr
library(MASS)
library(dplyr)

# force which select (if MASS was already loaded after dplyr and overwrite the function)
select <- dplyr::select

# or unload MASS
detach("package:MASS", unload=TRUE)

# check to see
environment(select) # if "MASS", you're in for a bad time
</code>
===== Backlinks =====


{{backlinks>.}}