Bear Carlson - bcarlson1@ufl.edu
git clone https://github.com/bear-is-asleep/PAD2.git
cd PAD2
source init.sh # (or source setup.sh if env is already setup)
python run_pad.py
Pad is a commissioning tool designed for SBND's PDS system. It's main purpose is to verify the channel mapping of the PDS components by overlying the light the observe with low level reconstructed tracks. The two types of tracks we use in commissioning are muon tracks and CRT tracks. When a particle passes through the TPC, the PDS components closest to the track are expected to see the most light in general. So PAD is able to view the cumulative light within a time window collected by each PDS. Additionally, clicking on a PDS when in the PAD window shows the raw waveform in the plot window below. PAD also supports both TPCs which are updated simultaneously. Lastly, PAD runs on dask which skips port forwarding on the browser which makes PAD much faster than using terminal port forwarding.
To install clone this repo wherever you want - I'd recommend your app directory
git clone https://github.com/bear-is-asleep/PAD2.git
reco1 and reco2 stages and are in development.
The data requires optical detector information but the waveforms are optional. First set up sbndcode
source /cvmfs/sbnd.opensciencegrid.org/products/sbnd/setup_sbnd.sh
setup sbndcode v09_73_00 -q e20:prof (earlier versions not supported by PAD)
Now you run your fcl in the typical way up to detector simulation -
lar -c <your-generation-fcl>.fcl
lar -c standard_g4_sbnd.fcl -s [Gen root file].root
lar -c standard_detsim_sbnd.fcl -s [G4 root file]
Waveforms come from the software trigger producer, so we need to run the pmt trigger chain
Then run the pmt trigger chain
lar -c run_pmttriggerproducer.fcl -s [DetSim root file]
lar -c run_pmtArtdaqFragmentProducer.fcl -s [PMT hardware trigger root file]
Next make sure you set SaveHists: true in the pmt software trigger producer fcl to produce test_hist.root which contains the waveforms. Now you can get the waveforms by running
lar -c run_pmtsoftwaretriggerproducer.fcl -s [PMT fragment simulation root file]
Hitdumper is the commissioning tree that contains all of the available commissioning data. It also contains truth neutrino energy from GENIE. You can run this directly on a detsim file or on the software trigger producer output file. If you want software metrics make sure this line in the hitdumper fcl is set to true readpmtSoftTrigger: true. If you want PAD to show the full reconstructed PE (deconvolution, noise filtering, etc.) set this line to true readOpHits: true.
The commissioning muon tracks are not turned on by default, but are optional for pad. If you want them in hitdumper set this line to true readMuonTracks: true and this line to true readMuonHits: true.
If you want these from hitdumper set readMCParticle: true
You must run reco1 and reco2 stages to get this information
Now that the fcl is prepared you can run
lar -c run_hitdumper.fcl -s [DetSim or PMT software trigger root file]
To initialize the python enviroment - only need to do this once
source init.sh
To set it up just requires setting up the enviroment again
source setup.sh
You can also make your own configuration in config/ and import it here.
Loader drives PAD and sets up all of the products to include. You can find out more about the parameters of the loader by running python run_pad.py --help
python run_pad.py --help
usage: run_pad.py [-h] [--data DATA] [--hdump_name HDUMP_NAME] [--sm_name SM_NAME]
[--wfm_name WFM_NAME] [--load_muon LOAD_MUON] [--load_crt LOAD_CRT]
[--load_mcpart LOAD_MCPART] [--mode MODE]
[--filter_primaries FILTER_PRIMARIES] [--crt_filter_tpc CRT_FILTER_TPC]
[--mcpart_filter_tpc MCPART_FILTER_TPC]
Load data - default values are stored in the config selected
optional arguments:
-h, --help show this help message and exit
--data DATA Path to the data
--hdump_name HDUMP_NAME
hitdumper file name
--sm_name SM_NAME pmt software metrics file name
--wfm_name WFM_NAME waveform file name
--load_muon LOAD_MUON
load muon info
--load_crt LOAD_CRT load crt info
--load_mcpart LOAD_MCPART
load mcpart info
--mode MODE Optical detector mode
--filter_primaries FILTER_PRIMARIES
Filter mcpart to +- 10us around beam window
--crt_filter_tpc CRT_FILTER_TPC
Filter CRT tracks to TPC
--mcpart_filter_tpc MCPART_FILTER_TPC
Filter MCPart to TPC
Alternatively you can set the filenames using a config. Set this line from config.default import * to point to you configuration file, which should be formatted like this
#default.py example -
#Get directories
DATA_DIR = '/sbnd/data/users/brindenc/PAD/test_fcl/v1' #Waveforms and hitdumper location
#Get fnames
HDUMP_NAME = 'hitdumper_tree.root' #HDUMP_NAME = None if you did not make hitdumper
WFM_NAME = 'test_hist.root' #WFM_NAME = None if you did not make waveforms
SM_NAME = 'test_hist.root' #SM_NAME = None if you did not make software metrics
PMT_ARA_NAME = 'maps/PMT_ARAPUCA_info.csv' #Sets channel id and locations
HDRKEYS = ['run','subrun','event'] #event id keys
#PDS settings
MMAX = 'dynamic' #Setting for max color. dynamic to set for every interval. global to set for max observable pe.
MODE = 'op' #op for full opreco, prompt for software pe prompt, prelim for software pe prelim
COATINGS = [0,1,2,3,4] #[undefined, coated pmt, uncoated pmt, vis xa, vuv xa]
T0_THRESHOLDS = [10.,1.] #Min pe to denote t0 [pmt,xa]
MAX_SPREAD = 1000 #Max spread of all pds's t0 in ns
t0 = -1600 #Start bin
t1 = 1600 #End bin for -1600 - 1600 ns in 2 ns steps
dt = 2 #ns step
#CRT settings
LOAD_CRT = True #CRT tracks
CRT_FILTER_TPC = True #Filter CRT tracks to ones just in TPC
#MUON settings
LOAD_MUON = True #Muon tracks
#MCPart settings
LOAD_MCPART = True #G4 primary particles
MCPART_FILTER_TIME = True #Filter MCPart to +- 10us around beam window
MCPART_FILTER_TPC = True #Filter MCPart to TPC
#PAD settings
VERBOSE = True
In the terminal at the base of your PAD directory type python run_pad.py <--options>. Dask will show a link to the event display like this
Dash is running on http://127.0.0.1:8050/
* Serving Flask app 'run_pad'
* Debug mode: off
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
* Running on http://127.0.0.1:8050
Press CTRL+C to quit
127.0.0.1 - - [21/Jul/2023 13:03:04] "GET / HTTP/1.1" 200 -
Opening http://127.0.0.1:8050 in a browser will take you to the PAD display.
Time sliders change the range to integrate the PE for each PDS component
- Moving the t0 slider will change the initial time for both TPCs
- Moving the window slider will change the final time for both TPCs
- Takes about 100 ms to update
A list of available runs are on the right
- Entering the run, subrun, event of an event from the list on the right and clicking submit will update PAD to that readout
- Takes about 10 s to update
The waveforms for each TPC are shown just below in ADC vs. time [us] (X-ARAPUCA waveforms are not included)
- Clicking on a PDS shows its waveform
- Takes about 50 ms to update
- Get X-ARAPUCA waveform information
- Get waveforms into a more suitable format without having to run the trigger chain.