Modeling on nested data frames
You'll be working on the US Army ANSUR II body measurement dataset, which has been pre-loaded as ansur_df
. The goal is to nest the data for both sexes so that you can simultaneously train two linear models, one for each sex. These models will derive a person's weight from their stature (height) and waist circumference. You'll then unnest the data to inspect the model's statistics produced by the glance()
function from the broom
package.
The dplyr
, broom
, and purrr
packages have been pre-loaded for you.
Side note: In the provided code, the purrr
package's map()
function applies functions on each nested data frame. Check out this package if you like using functions inside pipes!
This exercise is part of the course
Reshaping Data with tidyr
Exercise instructions
- Group the data by
sex
. - Nest the data.
- Unnest the
glanced
column.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
ansur_df %>%
# Group the data by sex
___ %>%
# Nest the data
___ %>%
mutate(
fit = map(data, function(df) lm(weight_kg ~ waist_circum_m + stature_m, data = df)),
glanced = map(fit, glance)
) %>%
# Unnest the glanced column
___