estimateWhiteningMatrix {BSL} | R Documentation |
This function estimates the Whitening matrix to be used in BSL with Warton's shrinkage and Whitening (“wBSL” method of Priddle et al. (2021)). The Whitening transformation and decorrelation methods are detailed in Kessy et al. (2018).
estimateWhiteningMatrix(
n,
model,
method = c("PCA", "ZCA", "Cholesky", "ZCA-cor", "PCA-cor"),
thetaPoint = NULL,
parallel = FALSE,
parallelArgs = NULL
)
n |
The number of model simulations to estimate the Whitening matrix. |
model |
A “MODEL” object generated with function
|
method |
The type of Whitening method to be used. The default is “PCA”. |
thetaPoint |
A point estimate of the parameter value with non-negligible posterior support. |
parallel |
A logical value indicating whether parallel computing should
be used for simulation and summary statistic evaluation. The default is
|
parallelArgs |
A list of additional arguments to pass into the
|
The estimated Whitening matrix.
Kessy A, Lewin A, Strimmer K (2018).
“Optimal Whitening and Decorrelation.”
The American Statistician, 72(4), 309–314.
doi: 10.1080/00031305.2016.1277159.
Priddle JW, Sisson SA, Frazier DT, Turner I, Drovandi C (2021).
“Efficient Bayesian Synthetic Likelihood with Whitening Transformations.”
Journal of Computational and Graphical Statistics (In Press).
https://arxiv.org/abs/1909.04857.
## Not run:
data(ma2)
model <- newModel(fnSim = ma2_sim, fnSum = ma2_sum, simArgs = ma2$sim_args, theta0 = ma2$start)
W <- estimateWhiteningMatrix(20000, model, method = "PCA", thetaPoint = c(0.6, 0.2))
## End(Not run)