Add nodes by edge: attempt at #294

[leoniedu]

Implement #294

[codecov]

Codecov Report

Attention: Patch coverage is 95.73171% with 7 lines in your changes missing coverage. Please review.

Project coverage is 94.28%. Comparing base (0980407) to head (d19ea03).

Files with missing lines	Patch %	Lines
R/match-points-by-edge.R	95.73%	7 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #296      +/-   ##
==========================================
+ Coverage   94.13%   94.28%   +0.15%     
==========================================
  Files          52       53       +1     
  Lines        6893     7057     +164     
==========================================
+ Hits         6489     6654     +165     
+ Misses        404      403       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[mpadge]

Thanks @leoniedu, but your commits have entirely reformatted the code, so I can't see what you've done. Can you revert the code back to what it was, and then ensure you implement your changes only. It would also still be good to have some example code to work with, so if you can, could you please paste that in the issue #294. Thank you!

[leoniedu]

Yes, it's more of a proof of concept, since I was having a hard time figuring out add_nodes_to_graph and how to adapt it for the by edge logic.

I have a couple of questions. Could you answer them, to help me get to the correct path? If this is not the appropriate channel to do so, let me know.

1. Why do we create edges from point to edge ends when (any (dmat [upper.tri (dmat)] < dist_tol)), even when intersections_only = TRUE?

1.1) Related, if the point is too close (i.e. < dist_tol) to any of the edge vertices, we shouldn't be splitting or creating vertices, right?

2. Is the order of the rows in the graph important? How?

3. Should we perhaps start by generalizing dodgr_insert_vertex and/or insert_one_edge so that it handles multiple x/y coordinates for the splits ? This could make the code more modular.