extremevalues: Detect outliers in one-dimensional data.

> library(extremevalues)

バージョン: 2.3.2

関数名	概略
`dpareto`	Pareto distribution
`evGui`	GUI to explore options and results of the "extremevalues" package
`extremevalues`	An R package for outlier detection
`getOutliers`	Detect outliers
`invErf`	Inverse error function
`outlierPlot`	Plot results of outlierdetection

dpareto / dpareto / qpareto / rpareto

パレート分布（所得の分布など）。特徴... 連続確率分布（離散型はジップ分布）

Arguments

xm
alpha
x
p
n

> qpareto(p = 0.5)
> rpareto(n = 50)

 [1]  1.630412  6.073130  1.249466  4.770578  2.811869  4.622061  2.204511
 [8]  1.891774  3.512302  1.024274  1.132239 16.453988  1.413683  1.079787
[15]  2.963708  4.007436  2.118197  1.078042  1.124064  1.014918  2.557590
[22]  1.847870  4.010750  3.256358  1.443233  1.567658  1.244257 25.228261
[29]  3.741187  1.172115  1.465387  2.781005  1.578223  1.121331  1.687921
[36]  1.317454  1.850014  1.852253  1.142024  8.386982  5.381176  4.318200
[43]  5.323968  6.323954  1.097773  1.359738  6.355245  1.565745  3.028133
[50]  1.026999

getOutliers

> getOutliers(y = rpareto(n = 50))

$mu
[1] 2.322622

$sigma
[1] 1.657416

$nFit
[1] 40

$R2
          [,1]
[1,] 0.8479894

$limit
     Left     Right 
-1.081295  5.726539 

$method
[1] "Method I"

$distribution
[1] "normal"

$iRight
[1]  1  2 15 36

$iLeft
integer(0)

$nOut
 Left Right 
    0     4 

$yMin
[1] 1.108513

$yMax
[1] 5.37689

$rho
 Left Right 
    1     1 

$Fmin
[1] 0.1

$Fmax
[1] 0.9

outlierPlot

外れ値のプロット

> y <- rlnorm(100)
> y <- c(0.1*min(y), y, 10 * max(y))
> K <- getOutliers(y,method = "I",distribution="lognormal")
> L <- getOutliers(y,method = "II",distribution="lognormal")
> 
> par(mfrow=c(1,2))
> outlierPlot(y, K, mode = "qq")
> outlierPlot(y,L,mode="residual")