# Explore Real Data

# Get data

Let's use ProtoDUNE-SP as an example. To look at raw data, you need to setup WCT with LArSoft first. Then, you can copy a data file using ifdh cp -D gsiftp://url/to/raw_data . to your local directory. Note that this requires you have set up the FNAL Kerberos authentication already. Otherwise, you can wget an example (~7GB) raw data file here (gsiftp link of the same file).

The raw data file has a special root format that contains larsoft objects. You can check the data structure of an event through the useful eventdump.fcl command:

$ lar -n1 --nskip 0 -c eventdump.fcl /path/to/raw_data.root

# Run signal processing

The following command performs Wire-Cell signal processing on the raw data using a fhicl configuration wcls-raw-to-sig.fcl:

lar -n1 -c pgrapher/experiment/pdsp/wcls-raw-to-sig.fcl /path/to/raw_data.root

There are two output files:

• output.root: A small file that only stores the deconvoluted signals in LArSoft objects. Eventdump shows its data structure:

$ lar -n1 --nskip 0 -c eventdump.fcl output.root
PROCESS NAME | MODULE LABEL. | PRODUCT INSTANCE NAME | DATA PRODUCT TYPE............ | .SIZE
DAQ......... | daq.......... | ContainerFELIX....... | std::vector<artdaq::Fragment> | ....?
DAQ......... | daq.......... | ContainerTPC......... | std::vector<artdaq::Fragment> | ....?
wclsraw2sig. | raw2sig...... | wiener............... | std::vector<recob::Wire>..... | 15360
wclsraw2sig. | tpcrawdecoder | daq.................. | std::vector<raw::RawDigit>... | ....?
wclsraw2sig. | raw2sig...... | gauss................ | std::vector<recob::Wire>..... | 15360

• protodune-data-check.root: A large file that records both the waveforms after noise filtering (hx_raw) and after deconvolution (hx_gauss) in TH2F. They can be looked at using simple ROOT scripts, or using the Magnify waveform display tool.

# 3D imaging (Experimental)

The 3D imaging code in WCT is still under development (algorithms are being ported from the Wire-Cell Prototype) and the performance may not be optimal yet. Nonetheless, 3D imaging can be performed on the previous deconvoluted signals as follows:

lar -n1 -c pgrapher/experiment/pdsp/wcls-sig-to-img.fcl output.root

The obtained 3D image are saved in a json format: clusters-apa?-0000.json, one per APA. You can later convert this format to fit the Bee 3D display with an independent python package wire-cell-python. Note that this package runs in a virtual environment. To setup, outside of the singularity container, do

sudo apt-get install python3-venv
git clone https://github.com/WireCell/wire-cell-python.git
python3 -m venv wcpy
source wcpy/bin/activate
pip install numpy vtk shapely
cd wire-cell-python
python setup.py develop # first time deployment, can ignore next time

Now you can merge the json files into a Bee format through a python script which you can find at wire-cell-python/test/wct-img-2-bee.py.

python wct-img-2-bee.py 'clusters-apa*.json'

Finally, you can upload the upload.zip file you just created to the Bee 3D display. For example, here is the 3D imaging result of the example event.

# Common issues and solutions

# I cannot get data using ifdh.

Make sure your Kerberos ticket didn't expire. Run kinit your-email@FNAL.GOV first.

# The fhicl or jsonnet file cannot be found.

The file may not be in your $FHICL_FILE_PATH or $WIRECELL_PATH. In particular, some of those files may only be in the /wcdo/src/wct/cfg directory when people develop. Adding the following two lines to your wcdo-local-myproj.rc file will help.

export WIRECELL_PATH=/wcdo/src/wct/cfg:$WIRECELL_PATH
export FHICL_FILE_PATH=$WIRECELL_PATH:$FHICL_FILE_PATH

find-fhicl(){
  fhicl_file=$1
  for path in `echo $FHICL_FILE_PATH  | sed -e 's/:/\n/g'`;do find $path -name "$fhicl_file"  2>/dev/null;done
}