localization_matrix#
- dageo.utils.localization_matrix(covariance, data_positions, shape, cov_type='lower')[source]#
Return a localization matrix
Build a localization matrix from a full covariance matrix based on specific data positions.
- Parameters:
- covariancendarray
The lower triangular covariance matrix
(nx*ny, nx*ny).- data_positionsndarray
Positions in the grid for each data point (e.g., wells), zero-indexed, of size
(nd, 2).- shapetuple
Dimensions of the grid
(nx, ny).- cov_type{‘lower’, ‘upper’, ‘full’}; default: ‘lower’
Matrix type of the provided covariance matrix.
- Returns:
- loc_matrixndarray
Localization matrix of shape
(nx, ny, nd).