How joeoliverevans/comptoncamera works
Project type and purpose
• This is a scientific-computing/research project (not a web or enterprise app). Its two pillars are:
a) Monte Carlo modelling of Compton scattering (“Monte Carlo” folder)
b) Image reconstruction from simulated/experimental data (“Image Reconstruction” folder)
• The README confirms: “Workspace for Compton Camera project. Will likely contain solutions for Monte Carlo modelling and Image Reconstruction.”
File structure
• Top-level directories:
• Image Reconstruction – data directories (Data_Raw → Data → Final Data), scripts for voxelization, geometry, reconstruction, and saved outputs.
• Monte Carlo – simulation scripts organized into “Old Sims” and “Python Work,” plus a main.py for orchestrating new runs.
• Under Image Reconstruction:
• Data_Raw – initial CSV/parquet dumps from simulations/experiments
• Data – cleaned/combined datasets in CSV and Parquet
• Final Data – curated, final Parquet datasets ready for voxelization
• SavedVoxelCubes – exported voxel-grid text files after PolarsToVoxels step
• Scripts – ErosionAndDialation.py, LoadingVoxelCube.py, PolarsToVoxels.py, RotationMatrix.py, Trying3DGeometry.py, ThickCones.py, VoxelstoRotation.py, plus helper modules (conesbyapex2, positiondistributions, eventpairspositiondistributions, etc.)
• main.py – orchestrator for the reconstruction pipeline
• Under Monte Carlo:
• Old Sims – prior/archived simulation variants
• Python Work – current design of scattering, data extraction, histogramming, variance reduction
• main.py – top-level entrypoint for new Monte Carlo runs
• There are also notebooks (totalerror.ipynb) and a TODO list.
README for additional insights
• The README is minimal but confirms the dual focus on Monte Carlo modelling and reconstruction.