edges table viewer: incorrect right coordinate displayed?

[savitakartik]

There seems to be a difference between the right coordinates displayed in the Edge table viewer vs the tree sequence edges table.

[savitakartik]

I'm not sure the correct coordinates are displayed in Edges plot either, so it could be that somehow the wrong coordinate is recorded in the edges_df. I'll dig into it!

[benjeffery]

I've not been able to replicate this in a couple of tree sequences I tried.

[savitakartik]

We've discused this already, but posting here so I don't forget: we think it's likely because in the edges_df we end up representing edges in terms of site IDs, instead of genomic coordinates.

[benjeffery]

I've not been able to replicate this in a couple of tree sequences I tried.

I was looking at the wrong thing and have replicated this now. It is very odd.

[benjeffery]

Of course! We aren't dequantising the tszip columns! I'll add some more tests here - I should have caught this in the preprocessing tests.