More coordinates per instruction

#1

Hello

I’m using the open source API to calculate a route between 2 points. However I would like to have more coordinate points returned per instruction, or more instructions… Is that possible? For instance, a coordinate point every 100 meter, or an instruction every 100 meter…

Thank you for the information
Bernd

#2

Those are two different concepts.

The instructions are carefully created to not be superfluous and at the same time can navigate correctly on the route at each needed road junction.

The coordinates are usually enough to have detailed route path (unless use simplification). For example a straight line doesn’t need more than 2 points, start and end.

#3

Thank you for the information. Maybe I should give you some details about what I want to achieve. It would like to create something where I can calculate/estimate where a car is on a given route. I know do this by taking the time for instruction and divide that by the points in an instruction. But if 1 instruction takes a long time, and has a small amount of points, then this location is hard(er) to calculate…

#4

Then the new path detail feature might be interesting for you. You can already use “time” and soon “distance”: https://github.com/graphhopper/graphhopper/issues/1305

#5

OH thank you, that looks fun to play around with…

#6

I have generated long distance bus routes with graphhopper which i am now merging. I have got many common route sections (10s of thousands!), but some have individual extra points, resulting in simple 1-point skipping short cut lines. e.g one line is ,1,1 1,2 1,3… and the other …1,1 1,3…
I havent sat down to play with graphhopper again to see why this is so, they are blatantly the same route and i can filter them out (only 2 points and makes a tiny detour lets drop it and merge into the 3 point option)
P.S. if anyone has millions of shape points that they want merging into common sections labelled with the array of shapes (bus routes) they have relate to . i can help ;).

#7

This is due to the path simplification removing unnecessary points. You can try the `edge_id` path detail so that you avoid removing the junction nodes AND could use the edge IDs for similarity identification