crogan / mmfe8_dataformats Goto Github PK

There are extraneous entries -can these be removed? They cause the octuplet analysis code to segfault :( @jrfarah
Thanks!

E.g. I start with raw scintillator / MM / TP dat files; I want a combined root file; what do I do?

@jrfarah

https://github.com/crogan/MMFE8_DataFormats/blob/master/src/combineTPGBT.C#L308

lol

It would make looping over / drawing hits easier. Suggestion: tpfit_n.

e.g.

tree->Draw("mxlocal", "tpfit_n==7 || tpfit_n==8")

or

for i in xrange(tree.tpfit_n): 
    print tree.tpfit_CH[i]

(moved from Alex's issue in trigger_daq)

@jrfarah

Minor request: could you rename the cloned tree to inform the downstream stuff that it's looking at a combined TP root file? Specifically, this expects it.

Could you add:

combdata->SetName("TPcomb_data");

after cloning?

raw2evtdat writes the first fifo timestamp to file:
https://github.com/crogan/MMFE8_DataFormats/blob/master/src/raw2evtdat.C#L145

raw2root writes the last fifo timestamp to file:
https://github.com/crogan/MMFE8_DataFormats/blob/master/src/raw2root.C#L144

I'll bug someone about this tomorrow.

I think this is a bug:

nhitscycle takes the number of hits received by the ADDC per BC. But the number of hits processed by the ADDC is limited to 8. So gbt_BCID can be filled too many times. Like, if there are 16 hits sent to the ADDC, the gbt_BCID branch will be filled 16 times, but it should only be filled 8 times.

Here's an example from 3550:

> f = ROOT.TFile.Open("mmtp_test_21.decode.root")
> tr = f.Get("GBT_data")
> tr.GetEntry(1)
> print len(tr.gbt_CH), len(tr.gbt_VMM), len(tr.gbt_MMFE8), len(tr.gbt_BCID)
59 59 59 72

I'll make a patch and a pull request.

combineTPGBT.py loops eternally at the end of its running.

Once it reaches the last fit packet (j == nfit), the GBT iterator (i += 1) never increments, because there is a continue just before incrementing. Mamma mia.