mesmer.stats.gaspari_cohn_correlation_matrices

mesmer.stats.gaspari_cohn_correlation_matrices(geodist, localisation_radii)

Gaspari-Cohn correlation matrices for a range of localisation radii

Parameters:

• geodist (xr.DataArray, np.ndarray) – 2D array of great circle distances. Calculated from e.g. geodist_exact.

• localisation_radii (iterable of float) – Localisation radii to test (in km)

Returns:

gaspari_cohn_correlation_matrices (dict[float : xr.DataArray, np.ndarray]) – Gaspari-Cohn correlation matrix (values) for each localisation radius (keys)

Notes

Values in localisation_radii should not exceed 10’000 km by much because it can lead to correlation matrices which are not positive semidefinite.

See also

gaspari_cohn, geodist_exact