Now that we require clade indices to correspond to the sorted index of an edge's child clade (#9 ), changing the reference sequence of a MAD becomes more complex. There are now two things which must be done to make this happen:
First, there's the (simple) modification to make edge mutations imply the correct compact genomes relative to the new reference sequence. This was previously the only thing that needed to be done to change the reference sequence of a MAD:
Next, we must update the order of the vectors of edges in NodeStorage.clades_ (which is a vector of vectors of edges, in which each vector of edges corresponds to a child clade) to match the new sorted order of that node's LeafSet with compact genomes relative to the new reference sequence.
Now that we require clade indices to correspond to the sorted index of an edge's child clade (#9 ), changing the reference sequence of a MAD becomes more complex. There are now two things which must be done to make this happen:
NodeStorage.clades_(which is a vector of vectors of edges, in which each vector of edges corresponds to a child clade) to match the new sorted order of that node'sLeafSetwith compact genomes relative to the new reference sequence.