time_varying_prior#

pymc_marketing.mmm.tvp.time_varying_prior(name, X, dims, X_mid=None, m=200, L=None, eta_lam=1, ls_mu=5, ls_sigma=5, cov_func=None)[source]#

Time varying prior, based on the Hilbert Space Gaussian Process (HSGP).

For more information see pymc.gp.HSGP.

Parameters:

name (str) – Name of the prior and associated variables.
X (1d array-like of int or float) – Time points.
X_mid (int or float) – Midpoint of the time points.
dims (tuple of str or str) – Dimensions of the prior. If a tuple, the first element is the name of the time dimension, and the second may be any other dimension, across which independent time varying priors for each coordinate are desired (e.g. channels).
m (int) – Number of basis functions.
L (int) – Extent of basis functions. Set this to reflect the expected range of in+out-of-sample data (considering that time-indices are zero-centered). Default is X_mid * 2 (identical to c=2 in HSGP).
eta_lam (float) – Exponential prior for the variance.
ls_mu (float) – Mean of the inverse gamma prior for the lengthscale.
ls_sigma (float) – Standard deviation of the inverse gamma prior for the lengthscale.
cov_func (pm.gp.cov.Covariance) – Covariance function.

Returns:

Time-varying prior.

Return type:

pt.TensorVariable

References

Ruitort-Mayol, G., and Anderson, M., and Solin, A., and Vehtari, A. (2022). Practical Hilbert Space Approximate Bayesian Gaussian Processes for Probabilistic Programming
Solin, A., Sarkka, S. (2019) Hilbert Space Methods for Reduced-Rank Gaussian Process Regression.