search

XENON1T / pax

The XENON1T raw data processor [deprecated]
BSD 3-Clause "New" or "Revised" License
16 stars 15 forks source link

adding coincidence levels in current pax versions #717

Closed feigaodm closed 6 years ago

feigaodm commented 6 years ago

This is to add a few variables for PMT coincidence analysis, copying @pdeperio 's fix in pax_v6.8.0.1.

feigaodm commented 6 years ago

Fixes https://github.com/XENON1T/pax/issues/718

tunnell commented 6 years ago

Uhh, there's an issue. How does this even work? One sec.

tunnell commented 6 years ago

tight_coincidence_thresholds = np.zeros(5, dtype=np.int16) # various tight coincidence levels

That's not how you define things in datastructure. Look at the other np arrays. You have to initialize the values in the plugins. You also want just the basic np.array type such that the ROOT code knows how to produce the ROOT C++ class.

feigaodm commented 6 years ago

I was trying to copy what was done for peak width variables, like 'range_area_decile = np.zeros(11, dtype=np.float)'

tunnell commented 6 years ago

Oh really? Odd. I thought we had very strict requirements on what datatypes could be converted. OK, it's too late right now, maybe bad to review.

tunnell commented 6 years ago

Due to #718, I would still try what I suggest. Look at area_per_channel for example. I'm not sure why the width code works, but this stuff is very finicky so it's worth trying to add the array like done for the others because that's clearly the source of the issue.