rrBLUP_ML0_neg2LogLik_fast

pybrops.model.gmod.rrBLUPModel0.rrBLUP_ML0_neg2LogLik_fast(logVarComp, etasq, d, n)

-2 log-likelihood function for ML estimation. In optimization, this function is to be minimized.

Parameters:

• logVarComp (numpy.ndarray) – A vector of shape (2,) containing parameter estimates in the log scale (log(varE),log(varU)). The log scale is used for more efficient search since the search space is (0,Inf).

• etasq (numpy.ndarray) – A vector of shape (ncomp,) containing squared eta values.

• d (numpy.ndarray) – A vector of shape (ncomp,) containing non-zero eigenvalues for the genomic relationship matrix.

• n (Integral) – Number of observations (nobs).

Returns:

out – Scalar value proportional to the -2 log-likelihood for ML estimation.

Return type:

Real