protectedCall: Wrap Function-Calls to Execute Additional Checks.

In VCA: Variance Component Analysis

Wrap Function-Calls to Execute Additional Checks.

Description

Function can be used to wrap function-calls, here, intended for model fitting functions anovaVCA, anovaMM, remlVCA, remlMM, fitVCA, and fitLMM. When wrapped, there is the option to perform additional checks and reporting back identified problems by setting 'ErrorType="Detailed"'. There is no error-handling provided by this function, i.e. any error issued will remain an error. It would need to be handled by try, tryCatch or similar. Note, that inline definition of datasets within 'expr' is not supported and will issue an error.

Usage

protectedCall(expr, ErrorType = c("Simple", "Detailed"))

Arguments

expr	(expression) to be protected, typically, a call to a model-fitting function from this package (see details)
ErrorType	(ErrorType) "Simple"=default error-messages, "Detailed"= additional data consistency checks will be performed

Author(s)

Andre Schuetzenmeister [e-mail]

Examples

## Not run: 
# nothing happens if no error occurs
data(dataEP05A2_1)
res <- protectedCall(anovaVCA(form=y~day/run, Data=dataEP05A2_1))
res
# error message without additional consistency checks (default)
dat3 <- data.frame(	y=rnorm(8,10),
						day=rep(c(1,2),c(4,4)), 
						run=rep(c(2,1), c(4,4)))
protectedCall(anovaVCA(form=y~day/run, Data=dat3), ErrorType="Simple")
# error message with additional consistency checks hopefully helpful for the user
protectedCall(anovaVCA(form=y~day/run, Data=dat3), ErrorType="Detailed")

# handle error
res <- try(protectedCall(anovaVCA(form=y~day/run, Data=dat3), ErrorType="Detailed"), silent=TRUE)
if(is(res, "try-error"))
	cat(sub(", ErrorType .*\\)", "", sub("protectedCall\\(", "", res)))

# inline-definition of data.frames issues an error
protectedCall(anovaVCA(	form=y~day/run, 
							Data=data.frame(y=rnorm(8,10),
								day=rep(c(1,2),c(4,4)), 
								run=rep(c(2,1), c(4,4)))))

## End(Not run)

VCA documentation built on May 29, 2024, 1:48 a.m.