Posterior predictive distribution

The weight_chains data frame (in your workspace) contains your 100,000 posterior predictions, Y_180, for the weight of a 180 cm tall adult:

> head(weight_chains, 2)
          a        b        s iter    m_180    Y_180
1 -113.9029 1.072505 8.772007    1 79.14803 71.65811
2 -115.0644 1.077914 8.986393    2 78.96014 75.78893

You will use these 100,000 predictions to approximate the posterior predictive distribution for the weight of a 180 cm tall adult. The bdims data are in your workspace.

This exercise is part of the course

Bayesian Modeling with RJAGS

View Course

Exercise instructions

Use the 10,000 Y_180 values to construct a 95% posterior credible interval for the weight of a 180 cm tall adult.
Construct a density plot of your 100,000 posterior plausible predictions.
Construct a scatterplot of the wgt vs hgt data in bdims.
- Use geom_abline() to superimpose the posterior regression trend.
- Use geom_segment() to superimpose a vertical line at a hgt of 180 that represents the lower & upper limits (y and yend) of ci_180.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Construct a posterior credible interval for the prediction
ci_180 <- quantile(___, probs = c(___, ___))
ci_180

# Construct a density plot of the posterior predictions
ggplot(___, aes(x = ___)) + 
    geom_density() + 
    geom_vline(xintercept = ci_180, color = "red")

# Visualize the credible interval on a scatterplot of the data
ggplot(___, aes(x = ___, y = ___)) + 
    geom_point() + 
    geom_abline(intercept = mean(___), slope = mean(___), color = "red") + 
    geom_segment(x = 180, xend = 180, y = ___, yend = ___, color = "red")

Edit and Run Code