search

ara-software / AraRoot

ARA data access framework
2 stars 7 forks source link

Martin01021207 patch 1 #41

Closed martin01021207 closed 2 years ago

martin01021207 commented 2 years ago

40 Update AraQualCuts short waveform criteria

A function named hasTooFewSamples() was added in both files AraQualCuts.cxx and AraQualCuts.h, this function detects waveforms whose number of samples is less than 500 (too few samples).

MyoungchulK commented 2 years ago

Hi Martin!

Thank you for the update and sorry for the late comment. Time calibration step removes bad sample during the calibration. So, the number of samples for calibrated WF is not always '64 * number of IRS blocks' and it can be varied... I agree with removing the WF that has a few samples. But maybe good to use the number of IRS blocks or the number of samples before time calibration. In that case, we don't have to worry about the variation in number of calibrated WF samples by a time calibration step

MyoungchulK commented 2 years ago

This is what would be 'proper' number of blocks for each event in each run. If you can read out IRS block information from each event and compare with this number (or number from config files) you don't have to count each WF length

martin01021207 commented 2 years ago

Hi, Myoungchul,

Thank you so much for the comments. I sent you an email which included the discussion we had with Brian and Dr. Ilya Kravchenko. It says why we decided to add another function that simply deals with waveforms whose number of samples is less than 500. Please let me know if there’s anything that you are concerned about or think I should know.

MyoungchulK commented 2 years ago

Thank you! I will read through the email:)

clark2668 commented 2 years ago

For documentation purposes, the following histogram of A2 2013 data is what supports the decision to set the cut at 500 samples. It comes from Brian Clark's thesis, section 4.5.3.

Screen Shot 2022-09-07 at 7 56 53 AM

clark2668 commented 2 years ago

I think this generally looks good. Is there any further update re: calibrated vs uncalibrated traces? In A2/3, we use calibrated traces. It could probably be done for uncalibrated traces too, but someone would need to make that histogram.

Martin, should we add this to the isGoodEvent function too? Here: https://github.com/ara-software/AraRoot/blob/master/AraEvent/AraQualCuts.cxx#L57 I think so?

martin01021207 commented 2 years ago

@clark2668 Hi, Brian, Thank you! I've just updated the isGoodEvent function. I don't have any update for the calibrated or uncalibrated traces, but maybe this is something that we can also discuss with Dr. Kravchenko. @ikrav