LocationScaleEstimates

LocationScaleEstimates estimates mean, pseudomedian, sd, mad and mode (of density) of a numeric variable. Estimates and CIs are based on t.test and wilcox.test for mean and pseudomedian. For other parameters the CI is based on bca-bootstrap relying on package boot. CAUTION: Bootstrap may take a while with large data.

LocationScaleEstimates(METRIC, PARAM = c("Mean", "Pseudomedian", "Sd"),
  CONF.LEVEL = 0.95, DIGITS = 3, MODEADJ = 1, NBOOT = 999)

Arguments

METRIC

is a numeric variable
PARAM

=c("Mean","Pseudomedian","Sd") vector of parameter names. Choose from PARAM=c("Mean","Pseudomedian","Sd","Mad","Mode")
CONF.LEVEL

= 0.95 Confidence Level for confidence interval
DIGITS

=3 Number of digits for estimates
MODEADJ

adjust, default=1 for density for mode estimation
NBOOT

=999 number of bootstrap replicates

Value

data.frame with parameters as rows and estimate, lower and upper 95

Examples

METRIC<-c(rnorm(200),rnorm(100,2,1),10:17)
if (FALSE) {
LocationScaleEstimates(METRIC)
LocationScaleEstimates(METRIC,PARAM=c("Mean","Pseudomedian","Mode"),MODEADJ = .7)
}