dyb_ann
Artificial neural network for accidental background subtraction
##Installation
On PDSF, use the following command
>module load git
>git clone https://github.com/hrpan/dyb_ann
For personal computers, make sure that git is installed, then simply clone the repository
>git clone https://github.com/hrpan/dyb_ann
And you are good to go.
##Evaluating a file
To evaluate a file, use apply.C with the following format
root 'apply.C("weightfile","inputfile","TTree")'
Make sure that your TTree contains four TBranch with the following names
- ep (Prompt Energy)
- ed (Delay Energy)
- dt (Capture Time)
- dist (Prompt-Delay Distance)
(Note that the naming has to be exactly the same, otherwise it won't work.)
The output file contains a TTree named MLP with TBranch eval_kMLP containing the responses from the given weight file.
###An example
First we generate the response file using apply.C
root 'apply.C("weights/test.weights.xml","ibd.root","IBD")'
then we load ibd.root with ROOT
>root ibd.root
Finally we add eval_kMLP as friend for IBD
root [0] IBD->AddFriend("MLP","ibd_kMLP.root")
Now you can use eval_kMLP as if it is in the same TTree as IBD