gam-gate 0.3.4

Simulation for Medical Physics

This experiment is a work in progress. Even the name (gam-gate) is temporary and will be changed.

How to install (short version)

First create an environment (not mandatory but highly advised)

conda create --name gam_env python=3.8
conda activate gam_env

Then install the package gam-gate. The package gam-g4 is automatically downloaded.

pip install gam-gate

If you already installed the packages and want to upgrade to last version:

pip install gam-gate -U

Once installed, you can run all tests:

gam_gate_tests

All tests are in the folder here but some data (binary files) are stored, for technical reasons, in this git: https://gitlab.in2p3.fr/opengamgate/gam_tests_data

---

How to install (long version, for developers)

There are two repositories:

• https://github.com/OpenGATE/gam-gate contains the python library, some tests and inside the gam_g4 folder the cpp library, linked to Geant4
• https://gitlab.in2p3.fr/opengamgate/gam_tests_data contains some test data (needed lfs)

The gam_g4 lib is composed of two folders:

• The folder gam_g4/gam_g4/g4_bindings contains C++ source code that maps some Geant4 classes into a Python module.
• The folder gam_g4/gam_g4/gam_lib contains additional C++ classes that extends Geant4 functionalities (also mapped to Python).

At the end of the compilation process of gam-g4 a Python module is available, named gam_g4. It is ready to be used from the Python side via the gam-gate python module.

⚠️ This is still work in progress and will probably changes ...

⚠️ Folder and module names are with an underscore (gam_g4) while python package in pip are with a minus sign (gam-g4). Don't ask us why.

1) First, create a Python environment and activate it.

If you use conda:

conda create --name gam_env python=3.8
conda activate gam_env

See: https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html

If you dont use conda:

python3 -m venv gam_env 
source gam_env/bin/activate

See: https://docs.python.org/3/tutorial/venv.html

2) Install required softwares (Geant4 and ITK)

Install Geant4 , with Multithreading=ON, and QT visualization. QT should work when install in the python environment, for example with conda install qt.

Install ITK.

3) Clone the repository (with submodules!)

git clone --recurse-submodules https://github.com/OpenGATE/gam-gate.git

4) Compile the c++ part

Go to gam-gate/gam_g4 folder and:

pip install -e .

It will create some files, ready to be compiled, but will fail, it is "normal" (well, it is not really normal, but the current way. Of course it will be improved).

Then, go in the created folder and cmake :

cd build/temp.linux-x86_64-3.6

cmake -DGeant4_DIR=~/src/geant4/build-geant4.10.07 -DPYTHON_EXECUTABLE=~/src/py/miniconda3/envs/gam_env/bin/python -DPYTHON_INCLUDE_DIR=~/src/py/miniconda3/envs/gam_env/include -DPYTHON_LIBRARY=~/src/py/miniconda3/envs/gam/lib/libpython3.so -DITK_DIR=~/src/itk/build-v5.2.0 -DROOT_DIR=~/src/geant4/build-root -DCMAKE_CXX_FLAGS="-Wno-pedantic"  . 

Of course, replace all folders by yours.

Then compile:

make -j 12

5) On linux

Sometimes on Linux machine, you will need to add the following path to find dynamic library :

export LD_PRELOAD=~/src/geant4/build-geant4.10.07/BuildProducts/lib64/libG4processes.so:${LD_PRELOAD}

We dont know yet why this is required and are currently working to improve this.

6) Test

Start python: python and type import gam_g4. The first time, Geant4 data will be downloaded. You can now access some G4 functions in Python.

For example:

import gam_g4
a = gam_g4.G4ThreeVector(0)
print(a)

7) Install the gam-gate python module

Then install the module:

cd gam-gate
pip install -e .

Several python's modules will be downloaded and installed (numpy, itk, matplotlib, etc)

8) Tests

The tests are available in gam-gate/gam_tests/ but you can run all available tests with:

gam_gate_tests