hi blasters,
blastlib2 is now ready for general use for reconstruction. i just
checked in a couple of bug fixes in the interface between the linking
and track fitting sections, and now, things seem to work fine.
there is an example macro 'recon.C' in the run directory, which
analyzes the
test data from
http://blast.lns.mit.edu/software/reconTest/test1/event.coda
(renamed test1.coda). after that, you can try 'reconSelect.C', which
only displays the events which did not reconstruct, with alternative
parameters commented out.
results: in the ideal data set, the fitter missed 20 of the 5000
events, mostly because of stringent convergent requirements. with these
relaxed in reconSelect.C, only 3 events were not fitted. i am working
on improving these.
--chris
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST