CLI Guide¶

This guide documents the command‑line interface for pyGSK.

Overview¶

pyGSK provides several subcommands:

simulate — generate raw power data and quicklook images
sk-test — perform the basic Spectral Kurtosis (SK) test
sk-renorm-test — perform renormalized SK validation
sk-thresholds — compute thresholds for SK
sk-thresholds-sweep — sweep thresholds across a PFA range

All commands follow the pattern:

pygsk <command> [options]

Architecture¶

ASCII diagram:

CLI args
   ↓
Subcommand parser
   ↓
runtests.* (for SK-related commands)
   ↓
simulate()  ← contamination described by `contam={...}`
   ↓
core.prepare_sk_input()
   ↓
core.get_sk(), core.renorm_sk(), compute_sk_thresholds()
   ↓
plot.* (if --plot)

simulate¶

pygsk simulate --ns 5000 --nf 32 --N 64 --d 1.0 --mode drift

This produces raw power ns × nf and displays a quicklook.

Contamination is defined via:

--mode noise
--mode burst --burst-amp A --burst-frac F
--mode drift --drift-amp A --drift-width-frac W ...

sk-test¶

pygsk sk-test --M 128 --N 64 --pfa 1e-3 --plot

Automatically simulates raw power unless --precomputed is supplied.

sk-renorm-test¶

Similar to sk-test but computes both raw and renormalized SK panels.

sk-thresholds¶

Computes thresholds for given M, N, d, PFA.

sk-thresholds-sweep¶

Sweeps thresholds across a PFA interval and optionally plots detection curves.