override approximate! so it copies edges, nodes, and pointsnode before computing accumHat. will need to split the accumHat portion out to its own function, so that trainedrotation doesn't have to know how accumHat is computed for different model types, eg., directed vs undirected
it should also copy
edges
,nodes
, andpointsNode
?override
approximate!
so it copies edges, nodes, and pointsnode before computing accumHat. will need to split the accumHat portion out to its own function, so that trainedrotation doesn't have to know how accumHat is computed for different model types, eg., directed vs undirectedhttps://github.com/snotskie/EpistemicNetworkAnalysis.jl/blob/master/src/linearmodeling.jl#L325