pax ROOT TTree Flattening

README

--------------------------------------    
Title: flattenPaxRootTree.C
--------------------------------------  

--------------------------------------    
Purpose: C++ program for flattening ROOT output of XENON1T pax processor. 
--------------------------------------    

--------------------------------------    
Compilation:

	0) On xecluster, run 'latest_root', i.e. requires ROOT >= v5.34

	1) Generate source code according to pax ROOT output format:

		./MakeClass.sh <path to pax processed ROOT file>

	This also builds the executable for the next step.

	2) Flatten the trees:

		./flattenPaxRootTree -i <path to pax processed ROOT file> -o <name of output flattened ROOT file>
--------------------------------------    

--------------------------------------    
Warnings:
	
	- Must re-generate source code (step 1) for any change in pax output format or pax run-time configuration.

	- If batch processing many files, length of filename used for MakeClass.sh must be the same for all files (since TTree::MakeClass creates a fixed-sized character array 
for pax_info).

	- For Mac OS X, need to modify sed commands in MakeClass.sh, since '-i' requires an argument: (sed -i) -> (sed -i '')
--------------------------------------    

--------------------------------------    
Notes: This is a very primitive program, quickly written to test a ROOT-based analysis. Modifications should be made in flattenPaxRootTree_in.C, e.g.
	- Selecting peak types and detectors to include
	- Ordering principle (currently reverse ordered by area for all peak types)
	- Which branches to save (currently all except index variables)
	
	or just re-write this all within pax. 
--------------------------------------    

--------------------------------------
Analysis: For interactive analysis, modify ana_flat.C with the file name you wish to analyze, then run 'root ana_flat.C' and use Event->Scan(), Event->Draw() as usual, e.g.

	 Event->Draw("log10(s2_tpc_area/s1_tpc_area):s1_tpc_area", "s2_tpc_area > 150 && s2_tpc_area < 5000 && n_s2_tpc < 10 && s1_tpc_hit_time_mean < s2_tpc_hit_time_mean && s
1_tpc_area<30","colz")

For batch (e.g. event-loop) analysis, a user-friendly framework still needs to be developed.
--------------------------------------