Computes summary statistics for alpha distributed as Gamma(a, b).
Usage
compute_alpha_diagnostics(a, b)
Arguments
- a
Numeric; shape parameter of the Gamma prior (> 0).
- b
Numeric; rate parameter of the Gamma prior (> 0).
Value
A list with components: mean (expected value Escales::alpha = a/b),
sd (standard deviation), cv (coefficient of variation = 1/sqrt(a)),
median, and quantiles (named vector with q5, q25, q50, q75, q95).
Details
The coefficient of variation depends only on the shape parameter a,
making it a useful summary of prior uncertainty regardless of the mean.
