We should do this so each flocker considers all agents within distance when calculating the separate heading vector as it is done in the other frameworks, now it just look for one, which means the total time it takes in Flocking-large benchmark is lower than it should be
Also it actually seems to me that the NetLogo version is too different from the other ones, it should probably be reimplemented from scratch to be really legit
We should do this so each flocker considers all agents within distance when calculating the separate heading vector as it is done in the other frameworks, now it just look for one, which means the total time it takes in
Flocking-large
benchmark is lower than it should be