Scribd
Upload
299 views

Data Wrangling in R PDF

This document discusses data wrangling in R. It introduces popular R packages for data import (readr), transformation (dplyr, tidyr), visualization (ggplot2), and communication (rmarkdown). The presenter advocates for tidy and functional programming approaches. Common challenges in data wrangling like cleaning, analyzing, and preparing data for sharing are addressed. Examples demonstrate how to use R and packages like dplyr to transform raw data into a tidy format suitable for further analysis.

Uploaded by

Arun Kumar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
299 views

Data Wrangling in R PDF

This document discusses data wrangling in R. It introduces popular R packages for data import (readr), transformation (dplyr, tidyr), visualization (ggplot2), and communication (rmarkdown). The presenter advocates for tidy and functional programming approaches. Common challenges in data wrangling like cleaning, analyzing, and preparing data for sharing are addressed. Examples demonstrate how to use R and packages like dplyr to transform raw data into a tidy format suitable for further analysis.

Uploaded by

Arun Kumar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

library(tidyverse)

Data Wrangling in R
Ben Best [email protected]
MBON Pole to Pole Brazil Workshop
2018-08-07
slides: bit.ly/r-wrangle-for-p2p
Motivation
● MBON Pole to Pole: develop a “Community of Practice”
○ Best practices
○ Common tools
● Manipulate (ie “wrangle”) data to:
○ Check quality
○ Analyze
○ Visualize
○ Publish
■ eg OBIS DarwinCore slides: bit.ly/r-wrangle-for-p2p 2
How many of you have used these?
1. R
scientific programming language
2. RStudio
integrated development environment (IDE)
3. dplyr
R package for grammar of data manipulation
4. rmarkdown
authoring framework for data science
5. git
version control system
6. Github
web hosting service for git to collaborate slides: bit.ly/r-wrangle-for-p2p 3
For Big-Data Scientists, ‘Janitor Work’ Is Key Hurdle to Insights
nytimes.com/2014/08/18/technology/for-big-data-scientists-hurdle-to-insights-is-janitor-work.html

“Data scientists, according to


interviews and expert estimates,
spend from 50 to 80 percent of
their time mired in the mundane
labor of collecting and preparing
data, before it can be explored
for useful information.”
–– NY Times (2014)
slides: bit.ly/r-wrangle-for-p2p 4
“Happy families are all alike; every unhappy
family is unhappy in its own way.”
–– Leo Tolstoy

“Tidy datasets are all alike, but every messy


dataset is messy in its own way.”
–– Hadley Wickham
slides: bit.ly/r-wrangle-for-p2p 5
Tidy Manifesto
cran.r-project.org/web/packages/tidyverse/vignettes/manifesto.html
1. Reuse existing data
structures

2. Compose simple
functions with the
pipe Fall into the
Pit of
3. Embrace functional
Success
programming

4. Design for humans blog.codinghorror.com/falling-into-the-pit-of-success

slides: bit.ly/r-wrangle-for-p2p 6
Tidyverse process & packages (unofficial in italics)
Program
Import Visualize
readr, readxl purrr
ggplot2
DBI glue
ggraph
jsonlite, httr rlang
ggmap
googledrive fs
htmlwidgets:
devtools, usethis
plotly
Tidy Transform leaflet Model
tibble dplyr dygraphs broom
dplyr forcats modelr
tidyr lubridate
Spatial hms
rgdal stringr
sp -> sf Communicate
raster knitr, rmarkdown
DT
shiny slides: bit.ly/r-wrangle-for-p2p 7
Which package do I use and how?
CRAN Task Views
cran.r-project.org/web/views

Cheat Sheets
rstudio.com/resources/cheatsheets

● RStudio IDE
● R Markdown
● Data Import (tidyr)
● Data Transformation (dplyr) slides: bit.ly/r-wrangle-for-p2p 8
Data Import cheat sheet
Tidy data organizes tabular data (as long) for use across R packages:

● Each observation is in its own row

● Each variable is in its own column

gather(data, key, value ) spread(data, key, value )


wide -> long long -> wide

slides: bit.ly/r-wrangle-for-p2p 9
Data Transformation cheat sheet
dplyr is a grammar of data manipulation, providing a consistent set of verbs that
help you solve the most common data manipulation challenges. (works w/ db’s)

● pipes, aka “then” operator %>% are elegant

read_csv('data/surveys.csv') %>% # read data


mutate(yr = year(obs_date)) %>% # add new column
select(species_id, yr) %>% # limit columns
filter(species_id == 'NL') %>% # limit rows
group_by(yr) %>% # group by values in column
summarise(n_obs = n()) %>% # then summarise by group
arrange(desc(n_obs)) %>% # sort in descending order
write_csv('data/surveys_clean.csv') # write out csv
slides: bit.ly/r-wrangle-for-p2p 10
Resources

r4ds.had.co.nz google.com stackoverflow.com


the “bible”
slides: bit.ly/r-wrangle-for-p2p 11
Demo(s)
rstudio::conf 2017 at Harry Potter World, FL

github.com/bbest/p2p-demo (TODO together)

Examples:
● bbest.github.io/ioos-bio-tidyr
● marinebon.github.io/info-intertidal
● marinebon.github.io/sbc-datasets
● bbest.github.io/bbest/p2p-demo-1
slides: bit.ly/r-wrangle-for-p2p 12

You might also like