from sam on slack:
"I think that, like with spatial, the non-spatial C2S and S2C should be dependent on the C2C - right now ,they are made separately and in general do not agree with one another, because they are using different sampling approaches (sampling for C2C, vs. all edges for C2S and S2C - not ideal)"
from sam on slack:
"I think that, like with spatial, the non-spatial C2S and S2C should be dependent on the C2C - right now ,they are made separately and in general do not agree with one another, because they are using different sampling approaches (sampling for C2C, vs. all edges for C2S and S2C - not ideal)"