time series versus runs (e.g. see Dario's list) -> Thibaut

[woohoudy]

Ok modification of lardon.py to extract value of pedestal into text files. Script routine to loop over different runs and then macro to read each text file.

[woohoudy]

Ok so first I extracted the Pedestal mean and RMS for raw data and filtered ones (4 variables). I used the data container of lardon as in "dc.evt_list[ievent].noise_filt.ped_mean" for extracting the value of the pedestal for filtered waveforms. I am always taking the first event of each run. The channel number is simply the index in the array "dc.evt_list[ievent].noise_filt.ped_mean". Is it after or before channel mapping??

Here you can see these 4 variables for run 435 as an illustration:

[woohoudy]

Then I look for each run. The run list is: Runlist.log

[woohoudy]

And finally one can look for evolution of these variables for the ~80 runs available. I selected some channels to make evolution of the RMS visible :

Same can be done for Pedestal value but it seems less interesting

[woohoudy]

Next step:

• Adding physical events to each run number (BE on, cosmic run, etc)
• Averaging over the first events for pedestal (even if in theory one event is enough no?)
• Playing with the new filters implemented to see if we got a better RMS
• Focusing on cosmic runs and fine tuning the display
• More?

[YoannKermaidic]

Updated RMS plots after fixing the FFT cut to 600 kHz (not 100 kHz - wrong previous setting) and after signal masking for the coherent noise treatment. The RMS is for single event only for now.

[woohoudy]

Comparing with Slavic's LarSoft Run 446:

To what is due the difference?? What is our mapping and his? DAQ channel? why do we have lower RMS on average but spikes?