McStas Stream Sampling to Nexus

Uses the StreamSampling.jl package to do weighted event sampling from McStas event data in NeXus format, and generate a NeXus TOFRaw event file.

Calling the StreamSampling Julia algorithm directly from Python (using juliacall) is about 30x faster than implementing a similar algorithm in Python.

Also included: a Dockerfile and an apptainer .def file to generate containers that have all the needed packages.

Requirements

• Julia installation and JuliaCall
  • Required Julia packages: StreamSampling, ChunkSplitters, Random, HDF5
  • After installing Julia, you can install them on the command line like so: julia -e 'using Pkg; Pkg.add(["StreamSampling", "ChunkSplitters", "Random", "HDF5"])'
• h5py, numpy, matplotlib
• scipp
• a json file describing metadata to save into the NeXus file
• McStas event file in NeXus format

Usage

python /path/to/sampling_to_nexus.py -h

for McStas files with multiple configurations: python /path/to/sample_multiconfig_mcstas.py -h

Note that for the multiconfiguration the configurations should be in the form /entry1/data/Config${confignum}_Panel${detnum}_event_signal_dat_list_p_x_y_n_id_t/events