1 min read

Using R with JupyterLab

2018/02/20

Jupyter Lab has been released. In addition to Python kernels, it can also run R kernels. Here is is how I set it up om my Macbook Pro.

  1. Install Jupyter Lab using instructions from the Readme.md file. I used
conda install -c conda-forge jupyterlab
jupyter serverextension enable --py jupyterlab --sys-prefix
  1. Install support for the R kernel. Instructions from Ben Chuanlong Du’s Blog were very helpful.

I started with this in R:

# install dependencies packages of IRKernel
install.packages(c('repr', 'IRdisplay', 'crayon', 'pbdZMQ', 'devtools'))
# install IRkernel
devtools::install_github('IRkernel/IRkernel')
# register IRkernel
IRkernel::installspec()

I then updated the configuration file:

jupyter lab --generate-config
  1. Finally, I tested a notebook with this R code chunk, adapted from hrbmstr. He appeared to be using Ride.

I modified his example like this. I needed to turn off warnings to stop useless font loading warnings.

# Need to turn off warnings for fonts
options(warn=-1)
library(ggplot2)
library(extrafont)
library(hrbrthemes)
loadfonts(quiet=TRUE)

plt <- ggplot (mtcars, aes(wt, mpg)) +
  geom_point() +
  theme_ipsum_rc(grid="XY")
print(plt)

options(warn=0)