BayesianSettings

BayesianSettings()

The arguments that change the posterior.

Attributes

Name	Type	Description
model_config		Pydantic configuration; frozen and closed to extra fields.
minimum_observations	int	Fewest error observations a track record needs before it yields a precision.
include_model	bool	Whether the forecaster’s own prediction enters the blend at all.
consensus_stdev_variance_multiplier	float	Scales the dispersion-based consensus variance.
consensus_count_exponent	float | None	Sharpens the dispersion-based consensus variance by the contributor count. None leaves it unsharpened.
consensus_bias_lambda	float	Fraction of the filtered consensus bias removed before blending. Zero leaves consensus unadjusted.
consensus_bias_process_noise	float	Process variance of the local-level filter that estimates consensus bias.
consensus_bias_measurement_noise	float	Measurement variance of that filter.
winsorize_fraction	float | None	Share of each tail clipped from historical growth and error inputs, each window on its own bounds. None clips nothing.
interval_width	float	The published interval’s coverage.
interval_df	float | None	Degrees of freedom of the Student-t interval. Omit all three shape parameters to use the calibration, which covers the dispersion-weighted profile at the calibrated bias lambda only.
interval_loc	float | None	Location of the Student-t interval.
interval_scale	float | None	Scale of the Student-t interval.
dispersion_weighted	bool	Whether consensus precision is tried from dispersion before falling back to its track record.
preset	str | None	Name of the preset these settings come from, kept when a preset is derived under a new version. None when built directly.
fitted_through	date | None	Last date the data behind any fitted value covered. None for chosen values. Filled from the calibration’s date when the interval comes from the calibration.
version	str	The identifier this settings object publishes under.