lme4GS (version 0.1)

theta_optim: Selection of bandwidth parameter for Gaussian and exponential kernels.

Description

Obtain the optimal value of the bandwidth parameter for the Gaussian and exponential kernels.

Usage

theta_optim(formula, data = NULL, Uvcov = NULL,
                       kernel = list(D = NULL, kernel_type = "gaussian", 
                                     theta_seq = NULL, MRK = NULL), 
                       verbose_lmer= 0L, verbose_grid_search=0L)

Value

A list that contains:

LL

Log-likelihood.

LL.max

Maximum of likelihood.

theta

Sequence of values for the bandwidth.

theta.max

Value of bandwidth when log-likelihood attains the maximum.

fm

Fitted model with the optimum bandwidth parameter.

K.opt

The kernel for the optimum bandwith parameter.

Arguments

formula

a two-sided linear formula object describing both the fixed-effects and random-effects part of the model, with the response on the left of a `~' operator and the terms, separated by `+' operators, on the right. Random-effects terms are distinguished by vertical bars (`|') separating expressions for design matrices from grouping factors.

data

an optional data frame containing the variables named in `formula'.

Uvcov

list.

kernel

list with the following elements, i)D: Distance matrix (can be NULL), ii) kernel_type: character, can be either "gaussian" or "exponential", ii)theta_seq: sequence of values for theta from which we select the optimum (can be NULL), iv) MRK: marker matrix from wich Euclidean distance is computed (can be NULL).

verbose_lmer

integer scalar, verbose output from optimizeLmer function?. If '> 0' verbose output is generated during the optimization of the parameter estimates, default value is 0L.

verbose_grid_search

integer scalar, if '>0' verbose output is generated, default value is 0L.

Author

Paulino Perez-Rodriguez, Diana Caamal-Pat

References

Caamal-Pat D., P. Perez-Rodriguez, J. Crossa, C. Velasco-Cruz, S. Perez-Elizalde, M. Vazquez-Pena. 2021. lme4GS: An R-Package for Genomic Selection. Front. Genet. 12:680569. doi: 10.3389/fgene.2021.680569 doi: 10.3389/fgene.2021.680569

Examples

	
# \donttest{
library(BGLR)
library(lme4GS)

data(wheat)

y = wheat.Y[,1]
X = wheat.X
A = wheat.A

rownames(X) <- rownames(A)

#model y=1*mu+Z_1*u_1+e, u_1~NM(0, \sigma_1*KG), KG: Gaussian kernel
wheat = data.frame(y=y, k_id=rownames(X))

fm1 <- theta_optim(y~(1|k_id), data = wheat, Uvcov = NULL,
                   kernel = list(D = NULL, kernel_type = "gaussian", 
                                 theta_seq = seq(3,8,length.out=10), MRK = X),
                   verbose_lmer=0L,verbose_grid_search=1L)

plot(fm1$theta,fm1$LL,xlab=expression(theta),ylab="Log-Likelihood")
fm1$theta.max
fm1$LL.max


# }