Dear INLA experts,
For seasonal latent effects, the periodicity of the seasonal effect is set with argument season.length. When you do this, does the periodicity depend on a unit of observation (data points)? For example, if I have a variable “date” which is measured year by year, as 2000, 2001, 2002… 2020, and I want to assume that the seasonal effect covers one year, then should I go as f(date, model=”seasonal”, season.length=1)? And should it be season.length=5, if I decide that the periodicity is 5 years?
Sincerely,
Rustam