create_m_and_L_recommendations#

pymc_marketing.mmm.hsgp.create_m_and_L_recommendations(X, X_mid, ls_lower=1.0, ls_upper=None, cov_func=CovFunc.ExpQuad)[source]#

Create recommendations for the number of basis functions based on the data.

Parameters:

Xnp.ndarray: The data.
X_midfloat: The mean of the data.
ls_lowerfloat: Lower bound for the lengthscale. Default is 1.0.
ls_upperfloat, optional: Upper bound for the lengthscale. Default is None.
cov_funcCovFunc: The covariance function. Default is CovFunc.ExpQuad.

Returns:

tuple[int, float]: The number of basis functions and the boundary of the approximation.