A copy of fax_waveform in python

[zhut19]

As suggested by @jhowl01, it's better leaving the original fax_waveform intact and make a separate folder for this alternative frame.

[JosephJHowlett]

This is an extension/reopening of #40, for the record.

[JosephJHowlett]

I think this should be merged - it will be useful for post-SR1 high-energy studies.

[tunnell]

Interesting. Is this to split fax outside of pax? Also, maybe @JelleAalbers interested since this may be useful for nT.

[JosephJHowlett]

To briefly review the history for posterity (to the extent I know it 😄) :

Qing built the set of scripts in this directory a couple years back, mainly since his study of reconstruction bias required large statistics of waveform-simulated S1s/S2s, without much regard for the spectra.

@pdeperio also used/uses some of the code in this dir in the full MC chain, to get the "truth info" from the fax instruction file into the same dataframe as the hax output.

@zhut19 made the code much simpler and removed all the bash calling, so now it is very simple to efficiently generate large statistics on compute nodes.

The framework is very useful for getting fax data if too much stats are required to directly run() pax in your notebook, and if you want to compare instructions to reconstructed information.

[tunnell]

Ah, but it's still fax, just wrapped a lot.

@jpienaar13 we should think for nT how all of this fits together between MC, strax, etc

[pdeperio]

@zhut19 @jhowl01 were you suggesting to change this in the full chain?

[pdeperio]

The output of fax truth info in full chain here is very slow in Pythonic analysis. This framework is a first step towards improving this, but will need to be revisited later in case of any more complicated truth-peak to reconstructed-peak matching analysis.