SR2 TFNN & FANN, Mar2019

XENON1T / pax

The XENON1T raw data processor [deprecated]

BSD 3-Clause "New" or "Revised" License

16 stars 17 forks source link

SR2 TFNN & FANN, Mar2019 #724

Closed weiyuehuan closed 5 years ago

weiyuehuan commented 5 years ago

one more PMT 71 off. All non-functional PMTs on top array: (1, 2, 26, 31, 62, 65, 71, 79, 87, 88, 91, 102, 118)

original and reconstructed TFNN position:

evt_dis

TFNN reconstructed error:

weiyuehuan commented 5 years ago

@feigaodm Do you know why this PR fails? seems python version problem? @mcfatelin please check which run_number we need to apply with PMT 71 off. hax setting

feigaodm commented 5 years ago

let's wait until PMT list is finalized. We don't want to change it further.

mcfatelin commented 5 years ago

I think in last Friday's computing meeting and in email exchange, PMT WG already confirmed about the final PMT list. See Julien's note. Maybe @ershockley @CaptainCooper @skazama can double check the PMT ids.

ershockley commented 5 years ago

Yeah I agree, my impression was PMT list is now finalized and we're just waiting on gain model to be implemented.

feigaodm commented 5 years ago

Can @weiyuehuan merge PR #726 to this? It will make it easier to track.

weiyuehuan commented 5 years ago

Thanks! @feigaodm @mcfatelin I've combined FANN and Tensorflow NN in this PR, and closed PR #726. let me see if I can find the problem, still can't pass the check..

feigaodm commented 5 years ago

oh, that's a known issue and couldn't be solved easily. It's not due to changes in this PR so you don't need to worry about it.