Distributions: part one

Visualization Best Practices in R

Nick Strayer

Instructor

What is distribution data?

Multiple 'observations'
Usually a sample of some population

Why distributions are important

Data collection or cleaning errors can become apparent
Could indicate the need to control for a variable in a model
Being true to the data

Standard plots

Histogram

Good for one distribution at a time
This chapter

boxplot

For comparing multiple distributions
Next chapter

Maryland speeding data

Speeding tickets given in Mongomery County, Maryland for 2017
Retrieved from data.montgomerycountymd.gov

md_speeding

Making a histogram in ggplot2

geom_histogram()
Automatically bins data for you
Just supply x aesthetic

md_speeding %>% 
  filter(vehicle_color == 'BLUE') %>% 
  ggplot(aes(x = speed)) +
  geom_histogram()

Let's make some histograms!

Visualization Best Practices in R

Preparing Video For Download...