post-workshop todos

[anastassiavybornova]

singletons:

all done

doubles:

• preserve attributes (CES logic) within remove_false_nodes, pass them to singleton solution so it doesn't get overwritten
• include loop where doubles which have C in the middle are treated (no code yet) - just take half of those, call singleton solution, recompute coins, solve other half
• combine it all in the end

clusters:

• first, solve non-plannar reconnections (xref reconsider non-planar edge reconnection to skeletonized artifact #139 )
• then, integrate function from n-clusters branch (clusters) into rest of workflow

workflow:

• we will have 3 functions (1, 2, many)
• wrap into single function that does split and calls in sequence
• in the end we want: singlefunction(streetnetwork) that solves face artifact detection, classification/typology, simplification, and returns finished network
• documentation for insiders

packaging:

• discuss with james

future feature requests:

• smoothening out of bumps
• dangling improvement
• optimally preserving attributes

[martinfleis]

Making a note for myself to look at this gap in Auckland

[jGaboardi]

I think this can be closed out. If not, we can reopen and transfer to sgeop.