Altering chains
Let's practice changing the number and length of chains so that we can get a posterior distribution of different sizes. By changing the size of the posterior, we can change the number of samples used for the posterior summaries, and impact the estimation time. The songs
data is already loaded.
This exercise is part of the course
Bayesian Regression Modeling with rstanarm
Exercise instructions
- For all models, predict
popularity
from thesong_age
. - Estimate a model with 3 chains, each 1000 iterations long, with the first 500 discarded
- Estimate a model with 2 chains, each 100 iterations long, discarding the first 50
- Print a summary of each model
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# 3 chains, 1000 iterations, 500 warmup
model_3chains <- stan_glm(___ ~ ___, data = songs,
___ = ___, ___ = ___, ___ = ___)
# Print a summary of model_3chains
___(model_3chains)
# 2 chains, 100 iterations, 50 warmup
model_2chains <- stan_glm(___ ~ ___, data = songs,
___ = ___, ___ = ___, ___ = ___)
# Print a summary of model_2chains
___(model_2chains)