Aliasing

class pyoptex.doe.fixed_structure.splitk_plot.metric.Aliasing(effects, alias, cov=None, W=None)

The sum of squares criterion for the weighted alias matrix. Computes the mean in case multiple Vinv are provided.

The effects indices from the model matrix are aliased against alias indices from the model matrix.

Attributes

covfunc(Y, X, Zs, Vinv, costs)

A function computing the covariate parameters and potential extra random effects.

Wnp.array(2d)

A potential weighting matrix for the elements in aliasing matrix A.

effectsnp.array(1d)

The indices of the effects in the model matrix to alias from.

aliasnp.array(1d)

The indices of the effects in the model matrix to alias to.

__init__(effects, alias, cov=None, W=None)

Creates the metric

Parameters

effectsnp.array(1d)

The indices of the effects in the model matrix to alias from.

aliasnp.array(1d)

The indices of the effects in the model matrix to alias to.

covfunc(Y, X)

The covariance function

Wnp.array(1d)

The weights for the aliasing matrix.

Methods

Aliasing.accepted(Y, X, params, update)	Updates the internal state when the updated design was accepted (and therefore better).
Aliasing.call(Y, X, params)	Computes the aliasing criterion.
Aliasing.init(Y, X, params)	Initializes the metric for each random initialization of the coordinate-exchange algorithm.
Aliasing.preinit(params)	Pre-initializes the metric
Aliasing.update(Y, X, params, update)	Computes the update to the metric according to update.