Skip to content

Latest commit

 

History

History
117 lines (72 loc) · 3.54 KB

README.md

File metadata and controls

117 lines (72 loc) · 3.54 KB
Raw

analysis code to reconstruct and analyze track/clusters from CYGNUS camera

Checkout instructions:

[email protected]:CYGNUS-RD/analysis.git git checkout tune_sel

Convert the H5 files in ROOT files with TH2D

This could be avoided reading directly the H5 files, but for now it is like this... See instructions in:

https://github.com/CYGNUS-RD/hdf2root

The usual name of the runs after the conversion is histogram_Run00494.root or histograms_Run01515.root if it was already a root file.

Updated HOW-TO-RUN

Running the analysis code (in general):

python3 reconstruction.py configFile.txt --pdir plots --max-entries X -jX -r R -t /inputdir/path/

  • configFile.txt is the configuration file with all the settings.
  • pdir is the directory where the plots will be saved.
  • max-entries is the number of images you want to analyse.
  • j is the number of cores you want to use.
  • R is the run number
  • /inputdir/path/ is the path to the directory where to put the input root files (for MC)

Running the analysis code on MC data (updated on June 2023):

Firstly, create pedestal map (example with pedestal run 4904 of LNGS):

  1. set justPedestal to True in configFile_LNGS.txt
  2. create pedmap with: python3 reconstruction.py configFile_LNGS.txt -r 4904

Now, you should have the file pedestals/pedmap_run4904_rebin1.root

Then, reconstruct digitized (simulated) images:

  1. set justPedestal to False (and debug_mode to 0) in configFile_MC.txt
  2. if needed, create the file pedestals/runlog_MC.csv with the script scripts/make_runlog_tmp.py. If the pedestal run is 4904, the simulated runs must be named with higher run numbers (4905, 4906 ...)
  3. move the simulated runs in the input directory /path/to/inputdir/ and recostruct them with:

python3 reconstruction.py configFile_MC.txt -r 4905 -t /path/to/inputdir/

Prerequisite to run:

  • Python 3.X (X>6)
  • Root 6.X

Python libraries:

  • cycler>=0.10.0
  • numpy >= 1.20
  • cython >= 3.0.2
  • cygnolib (see Install cygno-lib ) (which requires oidc-agent==4.2.6, boto3sts and midas== 0.0.1 )
  • scipy>=1.3.1
  • root-numpy==4.8.0 (only for older versions of the code before winter23 branch and with python < python3.10)
  • uproot==5.2.2
  • h5py==3.10.0
  • scikit_image==0.22.0
  • scikit-learn>=0.21.3
  • mahotas==1.4.13

(Beware that some dependent packages will be installed automatically as requirements for some of these packages like:

  • kiwisolver==1.1.0
  • matplotlib==3.1.1
  • networkx==2.4
  • pyparsing==2.4.2
  • python-dateutil==2.8.0
  • six==1.12.0

...so proceed in order)

Example

Download the code from github:

git clone [email protected]:CYGNUS-RD/analysis.git or git clone https://github.com/CYGNUS-RD/analysis.git

cd analysis

Get a file for a specific run taken with the DAQ (eg. run 2113):

wget https://swift.cloud.infn.it:8080/v1/AUTH_1e60fe39fba04701aa5ffc0b97871ed8/Cygnus/Data/LAB/histograms_Run02113.root

Change the run number in the config file (Line 34)

https://github.com/CYGNUS-RD/analysis/blob/fng_18/configFile.txt#L34

emacs -nw configFile.txt

Then run the code on all the events

python reconstruction.py configFile.txt

If your computer has X cores

(check on linux with cat /proc/cpuinfo | awk '/^processor/{print $3}’)

You can speed up the processing by parallelizing it:

python reconstruction.py configFile.txt -j X

You can now look at the output ROOT file with a tree containing 1 event/image with:

root -l reco_run02113_3D.root