Models with covariates

[Tess-LaCoil]

Implement a logistic regression on covariate data.
mu = linear model
p = logistic(mu)
growth rate = p * g(Y(t), theta)
so the covariates become an effect that moderates the maximum potential growth. There may be identifiability issues.

• Need to find out if Stan can have a loop to implement priors for each parameter
• Interactions get implemented outside the model maybe from a formula object?
• User passes in a table of covariates at the requisite levels.
• Do we allow year effects or just individual? May need different models for these.