checkGxM: Singularity Check for GxM Model Fittings

In GxM: Maximum Likelihood Estimation for Gene-by-Measured Environment Interaction Models

Description

This function is used to check the fitted models. Model fittings based on numerical integration and derivative-free optimization may encounter circumstances in which likelihood reaches infinity because of singular covariance matrix or other issues. This function helps to detect such circumstances.

Usage

checkGxM(GxMmle, dataset, rGvalue=0.995,
         localpercentage=2, localrange=20, outlyingextent=2)

Arguments

GxMmle	a model fitting outcome from running GxM.
dataset	a data frame which is used to produce GxMmle.
rGvalue	rGvalue which is used to produce GxMmle.
localpercentage	a local percentage in terms of the value of parameter whose absolute value of gradient is the largest among all gradient components.
localrange	a positive number of parameter points with which the density values are computed and investigated.
outlyingextent	a threshold value measuring the outlying extent. A pair of twins/siblings would be identified if the variation of density values from two close set of parameters is larger than this threshold.

Value

locallikelihood	a vector of local log-likelihood values with only one varying parameter whose absolute value of gradient is the largest among all gradient components. The length of locallikelihood is equal to localrange.
localfMP	a data frame matrix of local density values corresponding to the largest gap of locallikelihood based on two items of localresponse and their difference. localresponse represents contribution from individual pairs of twins/siblings to the overall log-likelihood with the estimated parameters from GxMmle.
outlierID	an integer indicating the pair of twins/siblings contributing to the singularity. A returned value zero implies that no such pair of twins/siblings is found based on the current settings.

See Also

GxM

Examples

id <- 1:10;
rG <- c(rep(1,5),rep(0.5,5));
M1 <- rnorm(10);
M2 <- rnorm(10);
P1 <- rnorm(10);
P2 <- rnorm(10);
Data <- data.frame(id,rG,M1,M2,P1,P2);
output <- GxM(dataset = Data, modelname='Chol', closedform=TRUE);
checkGxM(output, dataset=Data);

GxM documentation built on May 2, 2019, 3:38 p.m.