SimulationEdge lenght (int) vs edge.shape.length (double)

[schaemar]

Possibly inconsistent values

[praveale]

Values from SimulationEdge and corresponding EdgeShape are different

[matalukes]

SimulationEdge lenght is calculated in roadmap-processing/roadmaptools/calculate_curvature.py in function get_distance_between_coords from original coordinates, then its saved to geojson file and from that file it is parsed to java memory.
However edge.shape.length is calculated in geographtools/util/DistanceUtil.java by function computeEuclideanDistance from coordinates, that are transformed by geographtools/util/Transformer.java class and then rounded.
That causes different values

[F-I-D-O]

We now have to decide how to deal with the discrepancy.

[marekcuchy]

I think that the distance used for all the calculations should be taken directly from the graph source and it should not be calculated again from the coordinates which basically force us to use precise EdgeShapes even if we don't care about the shape at all because it is used by the core functionalities like Move activity. (I was currently forced to basically mock EdgeShape by extending it to return the length I needed).

Therefore, I suggest to remove the shapeLength parameter completely or at least force it to be equal with Edge.length and do not calculate it.