INLAPLUS

Github page to install the package is here.

1. Likelihoods: Gaussian, Poisson, Binomial, Beta

2. Latent Models: iid, RW1, RW2, Besag, bym2, or generic (any rankdef model for time or space)

3. Selection Criteria: LMLIK (log marginal likelihood), DIC

4. Startegies: GA, VBC (I have the code for SLA and LA but needs some scaling)

5. INLAPLUS can run using openmp and mpi.

1. CV (leave one out, leave group out) - Should be ready by Jan 30, 2023.

2. Allowing NA in the data (which is useful to predict missing data) - Should be ready by Jan 30, 2023.

1. When you have many of constraints, let's say more than 500?

2. When the size of the data doesn't exceed 30K - I will be working on moving INLAPLUS to GPU, hopefully the size issue will be solved.