Unadjusted Priors
Now let's specify a model that doesn't use adjusted scales for prior distributions, so that we alter rstanarm
default behavior. This will allow us to have more direct control over the information going into the estimation. The songs
data is already loaded.
This exercise is part of the course
Bayesian Regression Modeling with rstanarm
Exercise instructions
- Predict
popularity
fromsong_age
- Tell
rstanarm
not to autoscale the parameters - Print a prior summary to confirm there was no adjustment
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Estimate the model with unadjusted scales
no_scale <- stan_glm(popularity ~ song_age, data = songs,
___ = normal(___ = FALSE),
___ = normal(___ = FALSE),
___ = exponential(___ = FALSE)
)
# Print the prior summary
___(no_scale)