MARGARET

Modular Analysis and Reactor Graphics Application for Radiation and Energy Tallies

A PyQt6 application for loading and visualizing neutron flux data.

Features

• Pick a flux type (1D scalar, 1D time-dependent scalar; easy to extend) and declare the input dimension ordering (e.g. I x G vs G x I) so flux loads correctly however it was stored (I = spatial cell, G = energy group, T = time step).
• Load flux from .npy/.npz, .csv/.txt, and .h5/.hdf5. Multi-array containers prompt you to pick which array to load.
• Load an energy grid to label groups; generate or load the spatial and time grids. Time-dependent flux is scrubbed with a slider (y-axis held constant across time).
• Embedded matplotlib canvas with pan/zoom/save.

Download

Prebuilt executables for Windows, macOS, and Linux are attached to each GitHub Release (no Python install needed). Releases are built automatically when a v* tag is pushed; bump version in pyproject.toml to match when tagging.

• Windows: unzip and run margaret.exe. SmartScreen may warn because the exe is unsigned — click "More info" → "Run anyway".
• macOS: unzip and open margaret.app. The app is unsigned, so Gatekeeper blocks the first launch — right-click → Open, or run xattr -dr com.apple.quarantine margaret.app. Pick the arm64 zip for Apple Silicon or x64 for Intel.
• Linux: untar and run ./margaret.

The single-file executables self-extract on launch, so the first start takes a few seconds.

Install

pip install -e .          # runtime deps (PyQt6, matplotlib, numpy, h5py)
pip install -e ".[dev]"   # plus pytest / pytest-qt

Run

margaret              # installed console script (after pip install -e .)
python -m margaret    # run as a module

Layout

src/margaret/
  constants.py        app metadata + registries (FLUX_TYPES, FILE_FILTER, ...)
  core/flux_model.py  FluxModel - flux + grids + reductions (no Qt, unit-tested)
  io/arrays.py        file -> numpy array loaders + container introspection
  gui/plot_canvas.py  matplotlib-in-Qt canvas (1D line)
  gui/array_picker.py dialog to choose an array from a multi-array file
  gui/start_page.py   main window: builds UI, owns model + canvas, wires signals
  app.py              entry point (main())
  __main__.py         python -m margaret

Extending

• New flux type: add an entry to FLUX_TYPES in constants.py (its axes tuple drives the dimension-ordering options).
• New file format: handle its extension in io/arrays.py::load_array (and, for a multi-array container, in list_arrays).

Test

pytest          # core model + loader tests run headless (no display)