Survey-weighted density plots
We can also explore the shape of a variable with the smooth curve of a density plot. Let's create a density plot of nightly sleep where we facet by gender. Whereas the height of the histogram bars represent counts, the height of the density curve represents probabilities. Therefore, we will need to do some data wrangling before we create the plot.
This is a part of the course
“Analyzing Survey Data in R”
Exercise instructions
- For
NHANESraw
, remove rows that are missingSleepHrsNight
orGender
. - Grouping by
Gender
, add the columnWTMEC4YR_std
which equalsWTMEC4YR/sum(WTMEC4YR)
. - Pipe your wrangled data directly into a
ggplot()
whereSleepHrsNight
is mapped tox
andWTMEC4YR_std
is mapped to weight. Include a densitylayer
withbw = 0.6
andfill = "gold"
and a facetting layer where you facet byGender
.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Density plot of sleep faceted by gender
NHANESraw %>%
___(!is.na(___), !is.na(___)) %>%
group_by(___) %>%
mutate(WTMEC4YR_std = ___) %>%
ggplot(mapping = aes(x = ___, weight = ___)) +
geom____(bw = 0.6, fill = "gold") +
labs(x = "Hours of Sleep") +
____wrap(~___, labeller = "label_both")