In principle LogLikelihood fitting should also work when fitting to multiple data sets, however this SO question shows that it doesn't.
The issue seems to be that numpy as not able to take the log if the model components have different sizes. We should therefore take this log per component before summing.
In principle LogLikelihood fitting should also work when fitting to multiple data sets, however this SO question shows that it doesn't.
The issue seems to be that numpy as not able to take the log if the model components have different sizes. We should therefore take this log per component before summing.