Oke, this bug was solved. The patch will be included in the new version of Neurofitter I will release just before the Society for Neuroscience 2007 meeting.
Thx to report it.
Swarm Fitter incorrectly reports the best set of parameters it found. This can be traced to the implementation of it where on line 70 the results list is populated by the initial fly positions instead of the new fly positions.
