There are tons of resources out there for learning R. I've collated and categorised those that I've found useful and consider to be of high quality.

• Learning R

Details of the packages I will typically install are here along with how to update packages when you update R and how to list installed packages.

Packages you want/need will depend on you, but the CRAN Taskviews provide a useful overview to packages for particular tasks or areas of usage.

snippet.bash

> ascii_url <- "https://cran.r-project.org/src/contrib/Archive/ascii/ascii_2.1.tar.gz"
> install.packages(ascii_url, repos=NULL, type="source")

You'll have to either import data into R or generate data depending on whether you are performing analyses or simulations.

• Importing Data
• Simulating Data
• Tips & Tricks
• Reshaping Data
• dplyr
• stats19

There are many analyses that can be performed within R, and the number is growing rapidly as people write and make available extensions. Regardless there are some data manipulation and approaches to analysis that make life a lot easier.

• Missing Data
• Summary Tables
• Graphs with ggplot2
• Logistic Regression
• Linear Mixed Effects
• Generalised Linear Models
• Generalised Linear Mixed Models
• Survival Analysis
• Time Series
• Leaflet
• Model Diagnostics

Loops are actually quite slow in R, and where possible the problem should be cast using one of the apply() functions (there are several, mapply(), rapply(), sapply() and lapply()). Personally it took me some time to get my head around using these, and I found A Brief Introduction to apply in R invaluable. There are also other solutions such as some of the helper functions in the plyr package and in a lot of instances using the reshape2 package to melt() the data greatly facilitates the workflow.

More recently the purrr package provides a consistent and “tidy” approach to repeating tasks across lists under the guise of “functional programming”.

Because this is quite a large and varied topic I've split details of to a separate page

R integrates seamlessly with LaTeX using the wonderful Knitr package. Tips and tricks to facilitate common tasks are documented as and when I come across and resolve the problem.

More recently though I've switched to using RMarkdown, which is much more flexible allowing the production of HTML, LaTeX/PDF, M$-Word, and even integrating Shiny to produce dynamic/interactive web-pages.

• RMarkdown
• Knitr
• Shiny

Whilst essentially an output graphics is such a huge area it warrants its own section. There are many options for graphics in R, but they basically fall into two categories lattice graphics or ggplot2. I've opted to dedicate time learning the later, ggplot2 so there won't be much here on lattice.

• Stacked Bar Graphs
• Time Series Graphs
• Empricial CDF Graphs
• ggplot2 Extensions there are lots of them, here are some I find useful.
• ggplot2 Options there are tons of them, I always have to look them up, here are ones I use commonly.

Essentially any script written in R constitutes programming, but in this section I go into slightly greater detail about writing functions and keeping related functions grouped together as packages.

There is lots to R programming and unsurprisingly a lot has been written…

• Bookdown: Authoring Books with R Markdown by Yihui Xie
• R for Data Science by Garrett Grolemund and Hadley Wickham

If you're anything like me you'll regularly encounter error messages whilst working with R. I have attempted to curate those that I come across along with some of their meanings.

• Error Messages in R

A neat trick to update all installed packages whenever there is a major release of R is the following code…

snippet.bash

install.packages( 
    lib  = lib <- .libPaths()[1],
    pkgs = as.data.frame(installed.packages(lib), stringsAsFactors=FALSE)$Package,
    type = 'source'
)

Occasionally I've had packages where I've been provided updated versions and I need to install them manually rather than relying on CRAN (since I'm trying to install a version newer than on CRAN). This can be done with…

snippet.bash

install.packages('~/path/to/source-file-1.0.1.tar.gz',
                 repos = NULL,
                 type  = 'source')

The following was found on Reddit and is purportedly adapted from StackOverflow…

snippet.bash

list.of.packages <- c("assertr",
                      "ggplot2",
                      "tidyverse",
                      "magrittr",
                      "stringr",
                      "lubridate")
# Install missing packages
misssing.packages <- list.of.packages[!(list.of.packages %in% installed.packages()[,"Package"])]
if (length(misssing.packages) > 0) {install.packages(misssing.packages, dependencies = TRUE)}
# Load packages
lapply(list.of.packages, require, character.only = TRUE)
# Cleanup
rm(list.of.packages)
rm(misssing.packages)

R, being a statistical programming language, is really useful for statistical modelling.

A useful post on using XGBoost and SHAP for interpreting complex models is here

Shiny is an incredibly powerful tool for presenting your work/analyses.

• Shiny Server on Gentoo
• Shiny examples in your packages
• Shiny errors and their solutions
• Shiny Sample Size Calculator

A lot of people carry little computers around in their pockets these days (i.e. smartphones and tablets). Wouldn't it be great to have R in your pocket too? Well you can, and I've written up how to do this…Installing R on Android.

Another method using termux is detailed here.

Many people often state that they are worried about using R because its open-source and not validated. This is nonsense, all commercial products come with indemnity clauses that absolved the authors and publishers of any responsibility should the software be found to be faulty, thats Stata, SPSS, SAS and many more. The R Foundation have a document on compliance with the US's Food and Drug Adminstration (FDA) requirements R-FDA

The NIHR published Validation of Statistical Programming and the appendix includes an example of validating base commands in R in the appendix.

The question has cropped up on R-help many times (from memory, haven't time to search now I'm afraid).

If there are huge concern then its possible to use ValidR from Mango Solutions which provides a validated version of R. I would hazard a guess that many people are perfectly happy using Excel for work as its got a license but there are HUGE problems with it.

For the most part such fears are caused by people's ignorance and hopefully the above resources help educate and inform them to allay their concerns and encourage them to embrace open source software.

• R
• The R Journal
• R-help mailing list
• CRAN
• MRAN - Managed R-Archive Network
• R Task Views
• DataScience+ : R Tutorials
• Spreadsheet thinking vs database thinking

• jennybc/access-r-source: How to get at R source.
• Object Orientated Programming in R
• How to become a better R code detective? · Maëlle's R Blog

• tidymodels.org
• PCA and UMAP with tidymodels and #TidyTuesday cocktail recipes | Julia Silge

• R Documentation and manuals Search all 20,863 CRAN, Bioconductor and GitHub packages.
• All R documentation similar, slightly behind, allows you to run R in the browser.
• The R Graph Gallery – Help and inspiration for R charts
• Tidyverse Packages tend to have good documentation.
• CRAN Packages most packages have documentation in the form of a reference manual and many also have vignettes that accompany them that show how to use the package.
• R Graphics Cookbook

• BBC Visual and Data Journalism cookbook for R graphics
• Upset Plots - kieranhealy.org
• The Mockup Blog: Creating and using custom ggplot2 themes
• 30DCC (2022)
• Data Visualization Cheat Sheets | by Hubert Baniecki | ResponsibleML
• Graphic Design with ggplot2

• Coloring for Colorblindness
• Color and Contrast.com
• Palletes package ## Data Sets
• OpenFoodFacts.org

A curated list of books on many subjects can be found at Big Book of R.

• An Introduction to Statistical Learning With Applications in R (ISLR tidymodels Labs)
• The Elements of Statistical Learning: Data Mining, Inference, and Prediction
• R For Data Science by Hadley Wickham & Garrett Grolemund
• R Packages by Hadley Wickham
• Advanced R by Hadley Wickham
• Mastering Shiny
• Outstanding User Interfaces with Shiny
• Data Science with R by Garrett Grolemund
• Text Mining With R by Julia Silge and David Robinson
• Advanced Statistical Computing by Roger Peng
• Michael Clark: Statistics With R
• Hands-On Machine Learning with R
• Twitter for R programmers
• Thomas Mailund has written various books on R (and other things).
• Interpretable Machine Learning
• Tidy Modeling with R
• teacheR
• Time Series Analysis and Forecasting with ADAM
• R for applied epidemiology and public health | The Epidemiologist R Handbook
• Telling Stories With Data
• ggplot2 3rd Edition
• Introduction to Modern Statistics
• Applied Predictive Modeling - Kuhn & Johnson (topepo/tidy-apm: tidymodels code for the book Applied Predictive Modeling)
• The Hitchikers Guide to Responsible Machine Learning
• Introduction to Data Science
• R for Health Data Science
• Reproducible Analytical Pipelines - Master’s of Data Science
• Forecasting: Principles and Practice
• Marginal Effects
• What They Forgot to Teach You About R
• Introduction to Data Science
• Advanced Data Science

• Statistical rethinking with brms, ggplot2, and the tidyverse: Second edition (lecture notes rmcelreath/statrethinking_winter2019: Statistical Rethinking course at MPI-EVA from Dec 2018 through Feb 2019)
• Statistical Rethinking with brms, ggplot2, and the tidyverse: First edition * Stan User’s Guide

• CRAN Task View: Reproducible Research
• papaja
• Sharing - PLOS
• RMarkdown template checker - useful for checking RMarkdown Office templates before using them.
• R Workflow for Reproducible Biomedical Research Using Quarto
• The targets R Package User Manual

• Running a shiny app in a docker container | sellorm
• How to develop inside a Docker container to ease collaboration? - Rtask
• UseR!2022: Best Practices for Shiny Apps with Docker and More
• DevOps for Data Science

• QBit Workspace – Quantargo
• Serverless, On-Demand, Parametrised R Markdown Reports with AWS Lambda - mdneuzerling

• Bea Milz - Monitoring quarto-dev repositories: Creating a workflow with GitHub Actions for R users

You only really need to follow one blog to keep abreast of many people who blog about R….

• R-bloggers

Essentially audio blogs…

• Rtalk
• R-podcast
• SimplyStatistics not strictly about R but statistics more generally.
• Not So Standard Deviation again not strictly about R but covers RCatLadies so is worthy of inclusion.

I bookmark many articles on how to do things in R using Pocket, some are below.

• Scraping HTML tables using rvest
• Create a Glossary in R Markdown - Yongfu's Blog
• Android Smartphone Analysis in R \[Code + Video\] | R-bloggers
• Building A Neural Net from Scratch Using R - Part 1
• Building A Neural Net from Scratch Using R - Part 2
• Thread on learning R/Data Science
• Mapping iNaturalist Data in R
• Business Reports with R Markdown
• Read multiple files with filename as id

• The New Base Pipe | MyKo101
• Understanding the native R pipe |> | %>% dreams

statistics:R statistics:statistics statistics:programming