One of the packages I use all the time now in my interactive R sessions is pacman. It’s more than just a nicer way to load packages, it’s a great way to explore packages, functions, package libraries, versioning, etc.
Grab it off cran with
Out of the box, the function I use every day in all my scripts or at the console is
pacman::p_load(); it’s a great substitute for
require(). Yihui has already written about his preferences/understanding on the library vs. require.
pacman::p_load() for different reasons than he describes. Instead of having to write 5 lines of code to load for 5 common data munging packages, like so:
library(dplyr) library(tidyr) library(plyr) library(magrittr) library(ggplot2)
You can write one line
If you want to unload a package, say to update it, the function
pacman::p_unload() is pretty useful. Even better
pacman::p_load() will install and load new packages in the same command. And for github or bitbucket accounts, use
Yes, I know that if brevity in syntax for package loading were my main motivation, I could just do
import pandasfor python or
library(data.table)in R and get the equivalent functionality in one package load; but brevity of package loading isn’t my main concern in language or package choosing.
Something else that’s really nice is a way to print all the functions available in a package. Try it with
pacman::p_funs(magrittr). You may find the print out overwhelming for a large packages like
pacman::p_funs(stats); but I still like it in cases when IntelliSense auto-complete doesn’t quite get me to the function I’m looking for or is cramped for reading.
Also, whenever you’re doing a update to R or you just want to quickly back up all your packages, I find
pacman::p_lib() support helpful. For example,
Lastly, to find out what the path to your R-library, the function
pacman::p_path() is a real help. Check out the pacman github page for some more details.
In the next blog post, we’ll have a bit of fun with