metric.select.UI: Indicator metric selection

In p-schaefer/BenthicAnalysistesting: Benthic Analysis with Test Site Anlysis

Indicator metric selection

Description

Determines which indicator metrics which best differentiate the test site from its nearest-neighbbour reference sites. Metrics that indicate impairment will be used preferentially.

Usage

metric.select.UI(Test, Reference, outlier.rem = T, rank = F, outbound = 0.1)

Arguments

Test	Vector containing metric scores at the test site. Should be a single row from benth.met or add.met.
Reference	Data frame of metric scores at the reference sites. Should be output from benth.met or add.met.
outbound	Used if outlier.rem=T A numeric value between 0 and 1 indicating the outlier boundary for defining values as final outliers (default to 0.1)
Rank	Use rank differences in metric selection

Details

A interative selection algorithm is used as follows:

1. The first metric selected for the final set is the one which displayes the greatest distance from the Reference condition mean

2. Metrics with a pearson correlation greater than 0.7 to (any of) the selected metric(s) are excluded from further steps

3. The ranked departure of remaining metrics is divided by the (maximum) correlation with the metric(s) previously included in the analysis

4. The metric with the greatest score is selected for inclusion in the final set

5. Return to step 2 until the number of selected metrics is equal to the greater of 4 or 1/5 the number of Reference sites

If no metrics or too few metrics demonstrate impairment, the following metrics are included until the maximum is reached: Richness, Percent Dominance, HBI, Percent EPT.

Value

$Best.Metrics - Vector containing the final selected indicator metrics

$Indicative.Metrics - Vector containing all metrics that indicate impairment

$raw.data - Data frame containing only selected best metrics

$ref.sites - Vector containing input reference site names

$outlier.ref.sites - Vector containing sites removed as potential outliers

Examples

data(YKBioData,envir = environment())
bio.data<-benth.met(YKBioData,2,2)$Summary.Metrics
nn.refsites<- c("075-T-1", "019-T-1","003-T-1","076-T-1","071-T-1","022-T-1","074-T-1",
"002-T-1","004-T-1","073-T-1","186-T-1","062-T-1","005-T-1","025-T-1",
"187-T-1","023-T-1","193-T-1","192-T-1","196-T-1","194-T-1")
metric.select(bio.data[201,],bio.data[nn.refsites,])

p-schaefer/BenthicAnalysistesting documentation built on Jan. 17, 2024, 7:24 p.m.