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
It doesn't seem like the final messages being sent in the SimControl::finalize() function (namely the EntityPresentAtEnd messages) actually reach the SimpleCollisionMetrics plugin because run_network and run_metrics() are not called after publishing these messages. This causes incorrect output from the metrics plugin concerning whether an entity survived or not.
Furthermore, it seems as though some additional messages are getting hung up somewhere because after fixing this issue, the results presented by the SimpleCollisionsMetrics plugin (e.g. total flight time) seem to change (and look more accurate?). This can be seen when running the straight mission and looking at the metrics for the team with a single agent when that agent survives the entire mission; although the entire one agent team survives the whole simulation (100s) the total flight time is reported as substantially less (reported as 53 for me).
It doesn't seem like the final messages being sent in the SimControl::finalize() function (namely the EntityPresentAtEnd messages) actually reach the SimpleCollisionMetrics plugin because run_network and run_metrics() are not called after publishing these messages. This causes incorrect output from the metrics plugin concerning whether an entity survived or not.
Furthermore, it seems as though some additional messages are getting hung up somewhere because after fixing this issue, the results presented by the SimpleCollisionsMetrics plugin (e.g. total flight time) seem to change (and look more accurate?). This can be seen when running the straight mission and looking at the metrics for the team with a single agent when that agent survives the entire mission; although the entire one agent team survives the whole simulation (100s) the total flight time is reported as substantially less (reported as 53 for me).
Suggested a fix to this in PR #504
The text was updated successfully, but these errors were encountered: