Add in quadfitter #220
Add in quadfitter #220
Conversation
| @@ -0,0 +1,8 @@ | |||
| { | |||
| name: "FIT_QUAD", | |||
There was a problem hiding this comment.
This data entry should be placed in FITTER.ratdb, with the index of QUAD, rather than initing its own table.
There was a problem hiding this comment.
See my other comment. Happy to debate this.
Yes -- I'm working on that;) |
|
Hi @JamesJieranShen, thanks for the comments! I agree with all of them aside from the one about where to put the fitter ratdb table. Are we sure that we want to keep the tables for all fitters in the same file? I feel like that may become a bookkeeping nightmare, in SNO+ we allow each fitter to have it's own ratdb file. Each fitter might have a pretty large number of indices depending on the lifetime of the experiment and the complexity of the fitter. |
…g a check to make sure the point table size is larger than the maximum nhit value to use the table with.
Hey @smnaugle, I was referring to the json/ratdb internal data-structure. It should be in the same "table" as the other fitter, i.e. have the same "name", but different "index": {
"name": "Fitter",
"index": "QUAD",
...
}Onto splitting the fitters to different files -- I'm ok with putting the fitter ratdb entry as a separate file for book keeping purposes. However, clearly the current purpose of the FITTER.ratdb file is meant to include all fitter parameters. I would suggest changing the file of that name to FITTENSOR.ratdb if you would like to have a separate FITQUAD.ratdb. Thanks again! |
There was a problem hiding this comment.
LGTM, on the code front. I will leave to @MarcFBergevin and others to see if there's anything else wrong.
Adds in quadfitter as a rat processor that can be used by analyzers.
TODO: Should add in an experiment agnostic way for users to define what ratdb index to use for quadfitter.