Good practices with data

Keep raw data raw

Keep track of where the data comes from

No spaces, weird characters or symbols in file names or column names

Use a consistent folder structure

data
├── derived-data
│   └── 1-prep
│       └── cleaned-ed-visits.Rds
└── raw-data
    └── ED visits for Alec.xls

Good practices with R

Use RStudio projects

Use relative not absolute paths

Comment scripts throughout, keep track of changes

Use a common project structure

project/
├── data/
│   ├── derived-data/
│   └── raw-data/
├── graphics/
├── R/
├── epi-lunch-learn.Rproj
└── README.md

An example

Tidy data

• Packages are designed to work with tidy data (dplyr, ggplot2)
• Handles challenges like inconsistent number of diagnoses
• Data is unambiguous

Functions

Do you find yourself copy+pasting code, repeating lines on subsets of a data set, across different projects...

a2015 <- abs(mean(b2015) - b2015)
a2016 <- abs(mean(b2016) - b2016)
a2017 <- abs(mean(b2017) - b2015)
a2018 <- abs(mean(b2018) - b2018)

c2015 <- abs(mean(d2015) - d2015)
c2016 <- abs(mean(d2016) - d2016)
c2016 <- abs(mean(d2017) - d2015)
c2018 <- abs(mean(d2018) - d2018)

Problems...

• a greater risk of typos = hidden errors
• more lines of code = lost in your scripts
• more typing = tiring, carpal tunnel
• can't use in other projects or scripts = not reusable
• any change you make has to be made everywhere

Alternatively..

Write a function

calc <- function(b) {
  abs(mean(b) - b)
}
a2015 <- calc(b2015)
a2016 <- calc(b2016)
# ...

and apply that function over a list

lsYears <- list(b2015, b2016, b2017, b2018)
lapply(lsYears, calc)

Or across groups

bees %>% 
    group_by(year) %>% 
    calc(val)

Plotting

ggplot

geom_*

What kind of plot? (points, lines, histograms, ...)

aes

Links the data to aesthetic properties.

• ID -> color
• Treatment -> linetype
• value -> point size

ggplot(mtcars) + 
    geom_point(aes(mpg, cyl))

Spatial methods

• sf

Resources

• Advanced R
• Efficient R
• R for Data Science

Extras

• RMarkdown
• data.table
• GitHub