Tree structure for individual event analysis

ROOT
#1

Hi

I’m totally new to ROOT and I must say I’m a bit overwhelmed by the features and finally god totally lost :slight_smile:

I have to analyse several thousand events, each consisting of an array of times an ion hit a detector.
some thing like:
e1: [ 1, 3, 6, 8, 9, 10, 13, 17, 20]
e2: [ 2, 4, 6, 8, 9, 10, 12, 15, 17]
…

My question is how to save the events in the tree and get a histogram, fit, … per event?
Any hint is welcome.

Niko

#2

Hi,

welcome to ROOT :slight_smile: !
A good way to start is having a look to the ROOT primer: https://root.cern.ch/getting-started
For some “learning by doing”, you can have a look to the rich set of examples we provide: https://root.cern/doc/master/group__Tutorials.html

For your particular problem, you’ll need to parse your input file: what format are you dealing with? Can you already read the data in memory?
We can have a look to the rest of the steps once this is accomplished, it does not look like a hard task at all.

Cheers,
D

#3

Hi D,

We don’t have the DAQ at the moment, so I don’t have any “real” data yet. I like to think about how to handle the data first, and design the DAQ accordingly. Turned out this is easier than to adapt the post-proc to a weired DAQ.
What I propose is to have a raw-data-file with a flat structure starting with a short header with the trigger/event-count and UTC-time, followed by the relative event-times ending with some “DAQ-stop” line. For each sample-run I’ll have one file, consisting of thousands of such blocks.
Does that sound realistic?

It’s very nice to hear that you don’t consider this as big problem … to me it is … at least in ROOT :slight_smile:

cheers
Niko

#4

Hi,

I guess it is ok to write out a binary format if you discussed this within the DAQ community of your experiment and validated the procedure. I think the key here is to read these files in memory and store their content in some kind of C++ data tructures (e.g. std::vectors). This is the first step. Once this is accomplished, the ttree writing will be trivial.
Is the first step accomplished?

Cheers,
D

#5

Thanks

This is certainly manageable, we’ll make the DAQ in a way that vectors are available for post-processing.
I guess I’ll come back to this topic once we have some dummy data to tinker with.

cheers
Niko

#6

Hi Niko,

this sounds like a plan.

Cheers,
D

#7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.