train_CNBFA1: Train a copula NBFA1 model
In gregorp90/RStan-package-test: Copula-Based Factor Analysis

This function trains a copula factor analysis model with negative binomial marginals in Stan. Each group is assigned its own scaling parameter.

1	train_CNBFA1(train, gp_train, nfac = 2, ...)

`train`	data frame. The training data set, consisting of counts. Columns represent variables, rows represent observations.
`gp_train`	vector. Groups for each observation.
`nfac`	numeric. The number of factors.
`...`	Arguments passed to `rstan::sampling` (e.g. iter, chains).

The transformations are made separately, between the modeling of marginals and the correlation.

A list.

`train`	Training data set.
`gp_train`	A vector of groups for each observation.
`loadings`	Aggregated loadings.
`scores`	Aggregated factor scores.
`mu`	Aggregated mean parameters.
`phi`	Aggregated size parameters.
`Sigma`	Aggregated correlation matrix.
`stan_mod`	A list with two elements. Both are objects of S4 class `stanfit`. First is the marginal distribution estimation, second is the copula.