Tutorials

These tutorials are runnable Jupyter notebooks from the notebooks/ directory of the repository, rendered here with their saved outputs. Each one is self-contained. Download it and run it interactively to experiment. Notebooks 6–9 need slicersim and sncosmo installed (live noise).

Notebook	What it covers
1. Getting started	The observe → detect → plot loop, the SimulatedSpectrum fields, reproducible noise realizations, and reading the provenance in meta.
2. Atmosphere scenarios	The five scenario presets, overlaying them with .plot(ax=ax), why hazy can out-score clear, and the cold-planet temperature clamp.
3. Transits and binning	n_transits and R: the exact \(1/\sqrt{N}\) scaling, .plot_noise(), the instrument response and the two resolution limits.
4. Transit light curves	.light_curves().plot() for white-light and banded transits, validation of transit_flux against the analytic uniform-disk limit, limb darkening, and the trapezoid fallback.
5. Tables and targets	The table layout, index and sha256 manifest, the loaders, the frozen NEA registry, and where the caveats are recorded in the data itself.
6. The parametric model	The \(\zeta\) grid, the mean molecular weight from chemistry, the reconstruction of the POSEIDON tables, the validation table, and a custom sub-Neptune end to end.
7. A custom planet, end to end	A planet that is not in the registry (GJ 9827 d) from a dozen numbers; μ and deck variants; the same run through the forms/report web contract.
8. Atmosphere knobs	Sweeps of μ, metallicity, deck pressure, haze and \(T_\mathrm{eq}\) on a hot Jupiter and a sub-Neptune; feature amplitude, template S/N and transits needed.
9. Stars and noise	The PHOENIX SED grid, Vega→AB offsets, the error bar versus magnitude and band, the bright-host saturation floor, CBE vs EOL configurations.

Note

The repository also contains notebooks/verification.ipynb, a developer verification artifact that asserts physical sanity checks against the committed tables. It is deliberately not part of this tutorial series. Each notebooks/build_0N.py regenerates the corresponding notebook.