project8 / katydid

Project 8 data analysis package
Other
5 stars 4 forks source link

Update ROOT Writers with new discriminated point properties #136

Closed cclaessens closed 6 years ago

guiguem commented 6 years ago
guiguem commented 6 years ago

It seems that to be able to save Mean, NeighboorhoddAmplitude and Variance in TSparseWaterfallCandidateData, I will have to explode the TGraph2D into a tree.

guiguem commented 6 years ago

Looking into the MultiTrackEventData example, I will have to:

- [ ] Create a TClonesArray in the TSparseWaterfallCandidateData structure

@nsoblath Do you think this is correct or is there a better way?

nsoblath commented 6 years ago

I don't think you replace the TGraph2D with a TTree, but with a branch. That effectively gives you a sub-tree. I can look into this with you later.

cclaessens commented 6 years ago

I think we also want to add writing of sequential lines. So that optimization can look at all the intermediate processing steps.

guiguem commented 6 years ago

@cclaessens Can you write down the variables of the SequentialLineData you want to be saved in the ROOT file? It seems to me that SumX and so on are not interesting to have, and I would only save variable with names similar to the SparseWaterfallCandidateData so:

            MEMBERVARIABLE(unsigned, Component);
            MEMBERVARIABLE(uint64_t, AcquisitionID);
            MEMBERVARIABLE(unsigned, CandidateID);

            MEMBERVARIABLE(double, TimeInRunC);
            MEMBERVARIABLE(double, TimeLength);
            MEMBERVARIABLE(double, TimeInAcq);
            MEMBERVARIABLE(double, MinFrequency);
            MEMBERVARIABLE(double, MaxFrequency);
            MEMBERVARIABLE(double, FrequencyWidth);

Note that you are not creating the same variables there to, is there a need for having different variables output (like StartFrequency instead of MinFrequency)? Maybe we should unify what we can?

guiguem commented 6 years ago

Solved in 45d72130a3e1aa9ade3ef4f73b8263cc135f2bb2