Demonstrations

Every page in this section embeds the actual source file from the repository's demos/ directory — the documentation build pulls the scripts in directly, so what you read here is byte-identical to what you run.

All demos follow the same pattern: paths to SED_Tools data products are configured at the top of the file, synthetic "observations" are generated from known true parameters, and the recovery is checked against that truth — so each script doubles as an end-to-end validation you can run on your own grid and filters.

Demo	What it shows
Forward Model Demo	parameters → SED + synthetic magnitudes, with an annotated SED plot
Inverse Model Demo	magnitudes → posterior on (Teff, logg, [M/H]), with corner and trace plots
Parameter Modes Demo	the same data fitted under four fixed/bounded/free configurations
Extinction Comparison	one inverse fit per extinction prescription, compared against truth
Playground	a single configurable script for arbitrary forward + inverse experiments
API Usage Script	the minimal end-to-end workflow, suitable as a template

Requirements

The demos need the compiled Fortran extension (make), matplotlib (pip install "sed-model[demo]"), emcee for the inverse demos, and SED_Tools data products on disk. Each script's path block looks like:

GRID_DIR   = "~/SED_Tools/data/stellar_models/Kurucz2003all"
FILTER_DIR = "~/SED_Tools/data/filters/Generic/Johnson"
VEGA_SED   = "~/SED_Tools/data/stellar_models/vega_flam.csv"

Edit these to match your machine and run with python demos/<name>.py.