call

Aopt.call(Y, X, Zs, Vinv, costs)[source]

Computes the A-optimality criterion for a given design. Computes the average trace if multiple Vinv are provided.

Parameters

Ynp.array(2d): The design matrix
Xnp.array(2d): The model matrix
Zslist(np.array(1d)): The grouping matrices
Vinvnp.array(3d): The inverses of the multiple covariance matrices for each set of a-priori variance ratios.
costslist(np.array(1d), float, np.array(1d)): The list of different costs.

Returns

metricfloat: The negative of the A-optimality criterion.