I realized that when we changed the ResamplingVT object to return the mu and var by default (rather than the vt_factor), we got rid of the ability to enforce the n_eff > 4*n_obs condition. This PR adds a flag to include this check before returning the mu, var values in the call.
I realized that when we changed the ResamplingVT object to return the mu and var by default (rather than the vt_factor), we got rid of the ability to enforce the n_eff > 4*n_obs condition. This PR adds a flag to include this check before returning the mu, var values in the call.