MEDQObj-class: Class '"MEDQObj"'

In Blossom: Statistical Comparisons with Distance-Function Based Permutation Tests

Description

The MEDQObj class holds model specification and output from the Multivariate Medians and Distance Quantiles procedure. The summary.MEDQObj class represents the summary of these objects.

Objects from the Class

Objects can be created by calls of the form new("MEDQObj", ...) or, more commonly, via the medq function.

Slots

NumGrps:

The number of groups in the input data.

NumVars:

The number of response variables used in the multivariate medians and distance quantiles.

NumObs:

The total number of observations in the original data set.

GrpSizes:

A vector of length NumGrps specifying group sizes.

inputData:

A data.frame used for the analysis.

NumQuantVals:

The number of quantiles specified in the analysis.

quant:

A vector of quantiles used in the analysis.

MaxGpSize:

The number of observations in the largest group.

VariableWInGpMedian:

Within group median coordinates for specified variables.

GpAvgDistToGpMVMedian:

Average distance to multivariate median for each group.

GpMedQTolerance:

The solution tolerance.

WInGpQuantDist:

Within group distance quantile distance from the median.

ObsDistToGpMedian:

The observed distance to the multivariate median.

NumIterations:

The number of iterations for convergence to be reached.

Call:

The original function call.

group.names:

The names of the groups used in the analysis.

Methods

Dist2mvm

signature(x = "MEDQObj"): appends the distance to the multivariate median onto the the original input data and returns it.

print

signature(x = "MEDQObj"): same as summary

show

signature(object = "MEDQObj"): same as summary

summary

signature(object = "MEDQObj"): Prints a detailed summary of the MEDQ analysis.

Author(s)

Marian Talbert

See Also

medq

Examples

out <- medq(data = bgrouse)
summary(out)
#to get the data originally stored in the save option use
BgDist2mvm <- Dist2mvm(out)

Blossom documentation built on May 29, 2017, 10:55 p.m.