outInflu {nnDiag}R Documentation

Find Potential Outliers and Influential Observations

Description

Assesses the relationship between standardized residuals and the sum of times used as a neighbor. Outliers/influential observations will have a high standardized residual absolute value and/or have been used as a neighbor numerous times.

Usage

outInflu(object)

Arguments

object object of class "nnDgrps"

Value

An object of class "nnDoi", which is a list containing the following components:

neighbor.count number of times used as a neighbor in kNN classification
standardized.residuals standardized residuals (the ratios of residuals and their standard deviations)

Author(s)

Brian Walters walte137@msu.edu

References

McRoberts, R.E. (2009) Diagnostic tools for nearest neighbors techniques when used with satellite imagery, Remote Sensing of Environment. 113, 489–499.

See Also

grouper

Examples

data(LuceVolume)
data(LuceVolume_indx)

##First use grouper to make the "nnDgrps" object
grps <- grouper(LuceVolume$ref.volume, LuceVolume$pred.vol_k18, LuceVolume_indx)

x <- outInflu(object = grps)
## Not run: plot(x)


[Package nnDiag version 0.0-5 Index]