Developer Guide¶

This guide is intended for contributors extending pyGSK.

Package Layout¶

pygsk/
    core.py
    simulator.py
    thresholds.py
    runtests.py
    plot.py
    cli/
        main.py
        sk_cli.py
        sk_renorm_cli.py
        sk_thresholds_cli.py
        sk_thresholds_sweep_cli.py

Key Concepts¶

1. Canonical Data Path¶

All raw or simulated input must pass through:

core.prepare_sk_input()

This ensures consistent fields: - s1, s2 - time, freq - power (if available) - M, N, d

2. Contamination Model¶

simulate() now uses:

simulate(..., contam={mode:..., amp:..., ...})

CLI arguments are translated using:

_scrub_cli_kwargs()
_adapt_sim_cli_to_simulate()

3. Legacy Aliases¶

runtests uses _legacy_aliases() to guarantee plot compatibility.

4. Plot subsystem¶

Plotting is handled by:

plot_sk_histogram
plot_sk_dual_histogram
plot_data, plot_dyn, plot_lc

Lightcurve vs dynamic spectrum is selected automatically based on nf.

5. Developer Workflow (summary)¶

Work in a feature branch
Add tests in tests/
Ensure pytest passes
Update documentation
Submit PR to the upstream suncast-org repository