API Reference

The source already carries numpy-style docstrings (Parameters/Returns) for every public function, so help() gives a complete reference inline:

from varistar import TimeSeries

help(TimeSeries.mask_iqr_outliers)

Help on function mask_iqr_outliers in module varistar.timeseries:

mask_iqr_outliers(self, column: 'str | None' = None, error_column: 'str | None' = None, k: 'float' = 1.5) -> 'pl.Series'
    Build an IQR-based outlier mask on the fractional photometric error.

    Does **not** modify ``timeseries_df`` — pass the returned mask to
    ``apply_mask()`` to actually remove the flagged rows.

    Parameters
    ----------
    column : str | None
        Magnitude column (defaults to ``self.mag_col``).
    error_column : str | None
        Error column (defaults to ``self.err_col``).
    k : float
        IQR multiplier.  Standard box-plot uses 1.5; conservative = 3.0.

    Returns
    -------
    pl.Series[bool]
        True where the fractional error is an outlier.

Auto-generating this page

For a full reference without hand-maintaining it, consider quartodoc — it reads the same docstrings and generates one Quarto page per module/class, and slots into this same docusaurus-md pipeline. Illustrative config (run as a separate build step, not inside this notebook):

# _quartodoc.yml
quartodoc:
  package: varistar
  sections:
    - title: Core
      contents: [TimeSeries, LightCurve, TestGroup]
    - title: Period finding
      contents: [period.lomb_scargle, period.pdm, period.entropy]
    - title: Classification
      contents: [classify.variability, classify.eb_detector]

Module map

Module	Covers
varistar.timeseries	TimeSeries — I/O, cleaning, stats
varistar.lightcurve	LightCurve — period search, folding, fitting
varistar.groups	TestGroup — batch workflows
varistar.catalog	Survey loaders (OGLE, TESS, Gaia DR3, generic)
varistar.period	Lomb-Scargle, PDM, PDM2, SR, entropy/AoV
varistar.models	Fourier and Gaussian model functions
varistar.classify	Variability indices, EB detection
varistar.ml	Feature extraction, Dataset, FeaturePipeline
varistar.viz	Matplotlib styling, Plotly interactive plots