You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
See: cyclestreets/cyclestreets-r#86 for a new batch routing API parameter, successThreshold, which sets a success threshold for the percentage of routes successfully planned. This is part of work to move away from jobs getting stuck in finalising state.
When implemented in the package, I recommend you set it to lower than 99% for this project. You are sometimes dropping below that because of trying to plan routes too far from known cycle network, e.g. job 6801, so a level of 95% might be better.
The text was updated successfully, but these errors were encountered:
Great work. It would be worth documenting though in the code that this is not the same as the published API default. People could assume that 99% will be set without realising it.
You're getting a lower level about 95-98% because I think in this project you are planning an unusually high number of routes in areas with very sparse road networks, such that nearest point resolution is not finding a relevant start/finish point because there is none. Most other clients using it almost always get 100% with only very tiny numbers of failures when OSM has some dodgy data like dead-end one-way streets.
It may be better to set that at a per-project basis here at 90% but have the package use the API default of 99%. I think my suggestion above was rather unclear!
See: cyclestreets/cyclestreets-r#86 for a new batch routing API parameter,
successThreshold
, which sets a success threshold for the percentage of routes successfully planned. This is part of work to move away from jobs getting stuck in finalising state.When implemented in the package, I recommend you set it to lower than 99% for this project. You are sometimes dropping below that because of trying to plan routes too far from known cycle network, e.g. job 6801, so a level of 95% might be better.
The text was updated successfully, but these errors were encountered: