this is a big refactor that addresses a large bug, that allowed the variant names to get mismatched with their samples. basically this resulted in the wrong results being reported.
Most of this happens under the surface, with the addition of a posterior model that manages its own information and name, with several instances of it within the greater model. Additionally, any functionality that was based on a list is changed to a dictionary.
The only top level API impact from this is that the user no longer has to input variant_bucket_names at instantiation of the model, just the control bucket name, and the model does the rest
this is a big refactor that addresses a large bug, that allowed the variant names to get mismatched with their samples. basically this resulted in the wrong results being reported.
Most of this happens under the surface, with the addition of a posterior model that manages its own information and name, with several instances of it within the greater model. Additionally, any functionality that was based on a list is changed to a dictionary.
The only top level API impact from this is that the user no longer has to input
variant_bucket_namesat instantiation of the model, just the control bucket name, and the model does the rest