This is a scientific code originally developed to adress high speed - sparse data reconstruction in tomography.
The package contains support for reconstruction as well as generation of 4d phantoms for numerical experiments.
The central idea is to view the tomography problem as an inital value andvection PDE to be propagated in time. Solutions are obtianed by discretizing the density field in space in finite volume fashion and for each sought time recovering the driving velocity field through the projected advection equations using iterative methods.
NOTE
contomo is dependent on the astra-toolbox which needs a nvidia gpu with cuda to execute ray tracing.
To install contomo it is recomended to use Anaconda. Start in a new environment
conda create -n contomo python=3.8
conda activate contomo
Once in the new environment you will need to install pygalmesh(https://github.com/nschloe/pygalmesh)
conda install -c conda-forge pygalmesh
After this is is done, we go ahead and install astra-toolbox
conda install -c astra-toolbox astra-toolbox
Finally to install contomo clone the repo to your local machine
git clone https://github.com/AxelHenningsson/contomo.git
and next run a local installation in the repo folder as
cd contomo
pip install -e .
This will install some additional dependencies. You can try and run some things from the sandbox folder named test_... to see that things look ok.
Good luck!