MinnesotaPrior.sum_of_coefficients

MinnesotaPrior.sum_of_coefficients: float | None

Confidence in the sum-of-coefficients restriction \(\lambda_5\), or None.

When given, dummy_observations() contributes one artificial row per variable stating that a series at its pre-sample mean stays there, scaled by this value; larger imposes it harder. None contributes no rows. The same restriction is available as a separate SumOfCoefficientsPrior for composition with +; do not use both. Must be positive when given.