This is mostly an artifact of old GML graphs (e.g. the really big Shakya graph), but it's a good idea to get it taken care of regardless.
I guess, in the preprocessing script, if we identify a single node without edge weights given then we can just set some flag variable (and then a new column—let's call it has_edge_weights or something—in the database's assembly table) to FALSE.
This is mostly an artifact of old GML graphs (e.g. the really big Shakya graph), but it's a good idea to get it taken care of regardless.
I guess, in the preprocessing script, if we identify a single node without edge weights given then we can just set some flag variable (and then a new column—let's call it
has_edge_weights
or something—in the database'sassembly
table) toFALSE
.