R for data science

Getting started with Tidyverse R,
RStudio, and Quarto

John R Little

January 6, 2023

R for data science

Getting started with Tidyverse R, RStudio, and Quarto

John R Little
Center for Data & Visualization Sciences
Duke University

2023 January 06

Get code, data, and slides for today’s workshop
- https://github.com/libjohn/rfun_flipped
- https://github.com/libjohn/intro2r-code

15:00

Topics

  • How to use R
  • How reproducibility is easily accomplished
  • How to learn R efficiently
    • Part 1 (today): focus on data wrangling with dplyr
    • Part 2: visualization with ggplot2, briefly: EDA, interactive plots, linear regression models
    • Part 3: iterations & custom functions
    • Part 4: case study in wrangling and visualization by ingesting multiple excel worksheets from multiple excel workbooks

R, the Tidyverse, and Quarto

R is a programming language

  1. A data-first programing language → computational thinking
  2. The Tidyverse (and Tidymodels) is designed for humans
  3. Quarto Notebooks: a publishing system
    1. Publish high-quality articles, reports, presentations (slide-decks), websites, blogs, and books in HTML, PDF, MS Word, ePub, and more.
    2. Works with other languages and IDEs

Reproducibility

Reproducibility - Obtaining computational results using the same input data, computational steps, methods, code, and conditions of analysis. 1
Replicability - Obtaining consistent results across studies aimed at answering the same scientific question, each of which has obtained its own data.

Goals of a tool-based, first-class approach

  • Do as much as possible with code

  • Integrate prose with code; visualize inline

  • Generate reports for target audience

  • Iterate efficiently; {purrr}

Tidy data

  • Each variable is a column

  • Each observation is a row

  • Each type of observational unit is a table

Tidy data image from https://r4ds.hadley.nz/data-tidy.html#fig-tidy-structure

Image Credit: https://r4ds.hadley.nz

Citation: https://doi.org/10.18637/jss.v059.i10
Preprint: https://vita.had.co.nz/papers/tidy-data.pdf

See more in R for Data Science by Wickham and Grolemund

Tidyverse

  • A dialect of R

  • Easier to learn because of consistency and documentation

  • Assumptions

    • Data have semantic meaning that can be documented grammatically

    • Tidy data are wrangled, visualized, and iterated easily via grammar

    • 50-80% of any data project is data wrangling
      {dplyr} & {tidyr}

Pipe | data pipe | data sentence

 

A conjunction (“and then”), read left to right, creating a “data sentence”

 

%>% ({magrittr}|{tidyverse}) or |> (base R)

 

Starwars |>
    select(name, skin_color, homeworld)

Assignment

 

An object name “gets value from” a data pipe

 

<-

 

small_df <- starwars |>
     filter(gender == "feminine")

Project

  • Keep stuff organized in the same directory
    e.g. data, analysis, scripts, documentation, and outputs

  • In the notebook, refer to subdirectories via relative paths
    better than setwd()

  • Shareability, portability, legibility, and reproducibility
    Use Restart-R-and-Run instead of rm(list=ls())

Notebooks

Literate coding

  • Intersperse prose and code

  • Integrated outputs with analysis

  • Render reports from code

{dplyr}

library(dplyr) or library(tidyverse). Use {dplyr} to wrangle data

 

select

filter

arrange

mutate

group_by

summarize

subset by column

subset by row

sort

generate new variables


column totals (or subtotals with group_by)

Upcoming workshops

  • Visualization with ggplot2 (and interactive graphics) & Modeling (syntax)

  • Iteration and custom functions

  • Quarto and Observable interactivity

Rfun resources

  • Web scraping
  • Slide decks
  • Text mining, sentiment analysis, etc.
  • Dashboards ; interactivity
  • DBI: i.e. SQL without knowing SQL (working with databases)
  • git/GitHub

Next steps

Best way to learn and/or consultations

  • Take a small subset of a project you know well
    then recreate it in R
  • If you get stuck, schedule me for a free consultation
    walk-ins welcome
  • Documentation: {package-name}.tidyverse.org (https://dplyr.tidyverse.org)
  • Ask questions at RStudio Community ; R for DS online learning community ; R Ladies RTP
  • Formulate questions as REProducible EXamples (REPREX.tidyverse.org)

Fin

 

John R Little
Data Science Librarian
Center for Data & Visualization Sciences
Duke University Libraries
https://JohnLittle.info ● https://Rfun.library.duke.edu ● https://Library.duke.edu/data


Creative Commons: Attribution 4.0 ● https://creativecommons.org/licenses/by-nc/4.0

CC BY 4.0 John R Little