environmentaltools.spatiotemporal.bme.estimate_local_mean_bme

environmentaltools.spatiotemporal.bme.estimate_local_mean_bme(ck, ch, cs, zh, ms, vs, khs, order)[source]

Estimate local mean for Bayesian Maximum Entropy.

Computes local mean estimates at estimation points and data locations using polynomial regression weighted by covariance structure.

Parameters:

  • ck (pd.DataFrame) – Coordinates of estimation point with columns [‘x’, ‘y’, ‘t’].

  • ch (pd.DataFrame) – Coordinates of hard data locations with columns [‘x’, ‘y’, ‘t’].

  • cs (pd.DataFrame) – Coordinates of soft data locations with columns [‘x’, ‘y’, ‘t’].

  • zh (pd.DataFrame) – Values of hard data.

  • ms (pd.DataFrame) – Mean values of soft data.

  • vs (pd.DataFrame) – Variance values of soft data.

  • khs (np.ndarray) – Combined covariance matrix for hard and soft data.

  • order (list of int) – Polynomial regression orders [spatial_order, temporal_order].

Returns:

  • mkest (float) – Estimated mean at the estimation point.

  • mhest (np.ndarray) – Estimated means at hard data locations.

  • msest (np.ndarray) – Estimated means at soft data locations.

  • vkest (float) – Variance of the estimated mean at the estimation point.

Notes

Uses generalized least squares regression with covariance weighting to estimate spatiotemporal trends in the data.