fitNullModel doesn't converge when provided a list of matrices with more than 1 matrix

[tamartsi]

fitNullModel doesn't converge when provided a list of matrices with more than 1 matrix. We tried providing one matrix, and it was fine, we tried providing a list of a single matrix, and it was fine (converged). It seems like there is an issue with more than one matrix. I haven't looked at the code.

[smgogarten]

We have a unit test for multiple matrices, so it does run with a trivial example. Unfortunately that means the issue is more difficult to solve since it's related to the specifics of the data.

[tamartsi]

Hmmm. The fitNullModel function did work using an earlier version of GENESIS, so I don't think this is the data. Could it be related to the type of outcome? we looked at it today with a binary outcome (family = "binomial").

[mconomos]

Do you have a log you can share? It certainly has worked with multiple matrices in the past, and I don't think we've made any changes that would have affected this.