Open jGaboardi opened 1 month ago
If they are, no further simplification is needed, as far as I understand
wrong. Topological correctness does not mean that there are no artifacts.
There is nothing to do if the two data frames are equal and the artifacts dataframe is empty.
So that would be here after the first loop is completed, correct? (if that artifacts
df is empty)
When this condition is met, should it raise an error, or can we actually the input simplified enough and return? It seems like this could be true on both the first and second loops?
or can we actually the input simplified enough and return
We can't. Threshold detection occasionally fails in certain contexts due to the distribution of block shapes. It should raise as it does now.
OK, I'm wondering if it would be prudent to cut a patch release of momepy
after https://github.com/pysal/momepy/pull/666 so at least we can fail gracefully when an skeleton is passed in that can't be polygonized?
After that I can look back to try to determine better logic this.
We'd need to backport that. But if you wait a bit, I think that Streetscape will be ready for 0.9.
Yeah, I think this can surely wait. Just wanted to keep it on the radar.
We need some logic to end
sgeop.simplify_network()
immediately if the input roads dataset already meets our simplififed criteria. This can be determined within the call tosgeop.nodes.fix_topology()
and simply return the result there.The check should be after
induce_nodes()
is called where we are checking ifroads
androads_w_nodes
are equivalent. If they are, no further simplification is needed, as far as I understand.Thoughts @martinfleis @anastassiavybornova ?