gwinferno.models.bsplines.smoothing.apply_difference_prior

apply_difference_prior(coefs, inv_var, degree=1)

Computes the P-Spline difference penalty.

Parameters:

coefsarray_like

B-Spline coefficients.

inv_varfloat

Inverse of the penalty tuning parameter. Higher values result in smoother functions.

degreeint, default=1

Difference order.

Returns:

float

Log difference prior.