pyPDAF.PDAF.enkf_omega

pyPDAF.PDAF.enkf_omega()

Generation of a random matrix with orthogonal basis following SEEK approach for EnKF with given properties.

Parameters:

• seed (ndarray[tuple[4], np.intc]) – Seed for random number generation

• omega (ndarray[tuple[dim_ens, r], np.float64]) –

  Random matrix

  The1st-th dimension dim_ens is Ensemble size; the2nd-th dimension r is Approximated rank of covar matrix.

• norm (float) – Norm for ensemble transformation

• otype (int) – Type of Omega: (1) Simple Gaussian random matrix (2) Columns of unit norm (3) Columns of norm dim_ens^(-1/2) (4) Projection orthogonal (1,..,1)^T (6) Combination of 2 and 4 (7) Combination of 3 and 4 (8) Rows of sum 0 and variance 1

• screen (int) – Verbosity flag

Returns:

• omega (ndarray[tuple[dim_ens, r], np.float64]) – Random matrix

  The 1st-th dimension dim_ens is Ensemble size The 2nd-th dimension r is Approximated rank of covar matrix

• norm (float) – Norm for ensemble transformation