Closed guiguem closed 8 years ago
@guiguem: can you please also attach the configuration file you were using?
The data file is rid000001097-2016.09.13.22.55.32.411.MAT AlgorithmOptimization copy.txt
This should probably be handled in the writer, since a sparse waterfall plot with 0 points is a bogus track but still an important possible output of the track processing analysis. In this commit from last week on the Event Viewer branch, I simply had it throw a warning instead of trying to create the empty TGraph. Does this seem like a reasonable fix?
For what my opinion worths, that sounds reasonable... I added this piece of code in my branch, and fixed my problem...
Which piece of code? Can you push that change to GitHub? If it's fixing a bug that's relevant outside of your branch I will probably cherry pick it into a hotfix.
Mathieu is talking about adding these lines in KTROOTTreeTypeWriterEventAnalysis.cc around 300, before the TGraph2D is initialized:
if( points.size() == 0 )
{
KTWARN(publog, "The Sparse Waterfall Plot has no points! Will not write it");
return;
}
This has been fixed in the update I just made (commit 2ae94b1). It has been merged into develop and included in the release version 2.6.3.
After the two cuts on points during the track processing, if there is no more points, Katydid crashes when attempting to create a TGraph2D (see attachment). KTcrash.txt