Analysis processor framework for working with LArLite and LArCV data
larlitecv attempts to provide some tools to synchronize input from larlite and larcv data files. What exists now is a DataCoordinator class which allows one to get the right larlite and larcv data entry using run, subrun, and event values. Hopefully sometime in the near future, support for larlite ana_processor and larcv processmanager will be added.
larlitecv has the following required dependencies:
- gcc or clang
- ROOT (5 or 6, with 6 primarily supported at this point)
- larlite (https://github.com/larlight/larlite)
- larcv (https://github.com/LArbys/LArCV)
One must download and build the above before trying to compile larlitecv. Once everything is compiled, one must setup the environment for each each time a new shell is opened. There is a script, example_setup.sh, in the base folder of larlitecv to assist in this process.
Have compilers and stuff. Instructions on this is left to the internet. But on OS X install XCode and on Ubuntu install gcc and one should be pretty much set.
Install ROOT. This README will not cover how to do this. Please refer to ROOT documentation
Make a directory where all the magic will happen
$ mkdir work
Install larlite
$ git clone https://github.com/LArLight/larlite larlite
$ cd larlite
$ source config/setup.sh
$ make
[... if everything builds ...]
$ cd UserDev/BasicTool
$ make
$ cd ../../..
Install larcv. (Note shell must have larlite environment variables set. To set them up, run source config/setup.sh from larlite base directory. once you do this, the environment variable LARLITE_BASEDIR is defined.)
$ git clone https://github.com/LArCV larcv
$ cd larcv
$ source configure.sh
$ make
cd ..
Install larlitecv.
$ git clone https://github.com/larlitecv larlitecv
$ cd larlitecv
$ source configure.sh
$ make
cd ..
You should have the following directories in work
$ ls
larlite larcv larlitecv
[goto larlite folder]
$ source config/setup.sh
[goto larcv folder]
$ source configure.sh
[goto larlitecv folder]
$ source configure.sh
Alternatively, make a copy of example_setup.sh and change the MY_LARLITE and MY_LARCV to point to the right folders on your computer. Then run the script via
source my_example_setup.sh
$ root-config --libs
-L[path to root directory] -lCore -lRIO ...
$ larlite-config --includes
-I[path to larlite diretory]/core
$ larcv-config --includes
-I[path to larcv directory]/build/include -I[path to opencv include dir] -I[path to python include dir] ...
LArCV has an optional dependency on OpenCV, which gives access to some nice image processing functions. To use this, when compiling LArCV, one needs to define the following environment variables in the shell:
OPENCVLIBDIR
OPENCVINCDIR
These variables should point to the directory where the shared object libraries and headers for OpenCV live, respectively.
For example, on Ubuntu 16.10, this is
OPENCVLIBDIR=/usr/local/lib
OPENCVINCDIR=/usr/local/include
For those on OS X, one easy way is to install OpenCV using home brew or macports (under no circumstances should you be using both package managers! Pick one!). For home brew the above, you can find this info using brew info opencv3
There are two examples available in the app folder: Example and SelectionExample.
You can also make a minimial application using the command gen_larlitecv_app.py. For example,
$ gen_larlitecv_app.py test
This will make a new folder, app/test, which provides
- an empty class header,
test.h - an empty class source, `test.cxx
- an executable source file,
run_test.cxx - a GNUmakfile
- a LinkDef.h file (which helps ROOT build the symbols it needs to have your class be useable in the ROOT interpreter or callable in pyROOT)
You should be able to run make in this folder