Volatility

polars_ts.volatility

GARCH volatility modelling for time series. Closes #51.

GARCHResult dataclass

Fitted GARCH model result.

garch_fit(df, p=1, q=1, target_col='y', id_col='unique_id', time_col='ds', max_iter=200)

Fit a GARCH(p,q) model to each time series.

Parameters

df Input DataFrame (typically returns or residuals). p Number of lagged squared residuals (ARCH terms). q Number of lagged conditional variances (GARCH terms). target_col Column with the values to model. id_col Column identifying each time series. time_col Column with timestamps. max_iter Maximum optimization iterations.

Returns

dict Mapping from series ID to :class:GARCHResult.

garch_forecast(model, horizon)

Forecast conditional variance for horizon steps ahead.

Parameters

model A fitted :class:GARCHResult. horizon Number of steps to forecast.

Returns

list[float] Forecast conditional variances.