To install as a system library and executable "coresi"
pip install .
# or with the whell
pip install dist/coresi-0.0.X-py3-none-any.whl
conda env create -f conda-env.yaml
conda activate coresi
CORESI is known to work with pytorch version 2.4.1.
python3 -m coresi.main
# OR if coresi is installed
coresi
CORESI can run on GPU or CPU.
Make use the environment variable CUDA_VISIBLE_DEVICES
to tell which GPU
PyTorch should use.
The batch_scripts folder contains scripts that can be used to run CoReSi on the CREATIS and Jean Zay clusters.
CUDA_VISIBLE_DEVICES="1" python -m coresi.main
Supported flags:
coresi --help
usage: main.py [-h] [-v] [-c CONFIG] [--sensitivity] [--simulation] [--display]
CORESI - Code for Compton camera image reconstruction (default action)
options:
-h, --help show this help message and exit
-v, --verbose Enable debug output (default: False)
-c CONFIG, --config CONFIG
Path to the configuration file (default: config.yaml)
--sensitivity Compute the sensitivity and quits (default: False)
--simulation Do a simulation and quit (default: False)
--display Display the reconstructed image after the reconstruction (default: False)
See the config.yaml
file for configuring CORESI.
python -m coresi.display_image --help
usage: display_image.py [-h] -f FILE [-c CONFIG] [-s SLICE] [-a AXIS] [--cpp | --no-cpp]
CORESI - image display
options:
-h, --help show this help message and exit
-f FILE, --file FILE File path to the volume to display (default: None)
-c CONFIG, --config CONFIG
Path to the configuration file (default: config.yaml)
-s SLICE, --slice SLICE
Slice number (default: 0)
-a AXIS, --axis AXIS Axis x, y or z (default: z)
--cpp, --no-cpp Use this if the file comes from the C++ version of CORESI (default: None)
make tests
python3 -m build
This creates an installable wheel in the dist/
folder.