QC: comparison plots

Hi,

I would like to produce two sets of identical plots, for (forward) tracks : one set for MCH standalone tracks and one set for those MCH tracks that are in addition matched with MID.

So far in my dev version each set of plots is stored within a subpath of the mother task. So far so good. But what I’m really interested about are not so much the individual plot but the comparison of those plots.
What’s the best way to present such plots to the QC users ? I’ve thought about :

  • pick one of the set and for each histogram’s plot add to its list of objects the second histogram (e.g. in the beautify stage). I guess that works but then the second set stored in the second subpath is kind of redundant (and takes some space in the CCDB)
  • use the post processing to read both sets of plots and make comparison plots. At first it sounded like a good idea but looks a bit “heavy” (reading back the CCDB to get the two sets, while I have them both at hand in the first place, looks a waste of network bandwidth)

Any advice ?

Thanks,

Hi,

Are these plots produced by the same task and with the same samples ?
If so, you could keep the histograms in memory in the task but not publish them, while updating a third plot in EndOfCycle using these 2 other plots. This third plot would be published.
Not sure I have been clear.

Cheers,
Barth

Hi Barth,

Sorry for the late reaction. Yes, those plots are produced within the same task on the same input samples. So yes, I guess I could just publish the “merged” version updated at EndOfCycle. Will think about it.

Thanks,