think harder about how to set up latent inference code

goldingn / gpe

Gaussian process everything - an R package to reimplement and combine common statistical models using GPs

MIT License

26 stars 5 forks source link

think harder about how to set up latent inference code #10

Closed goldingn closed 9 years ago

goldingn commented 9 years ago

don't want to make predictions every time a model is fitted, so return marginal likelihood and components describing posterior at training (or inducing) locations

goldingn commented 9 years ago

This is done, marginal likelihood and some core components calculated and passed via a posterior object for predictions etc.