boxplot_cutpoints: Boxplot Outlier Cut Points

In ChrisDienes/SeaClass: Interactive GUI for analyzing a binary response variable

Description

Constructs boxplot style rules for flagging outliers.

Usage

boxplot_cutpoints(x, iqr_mult = 3)

Arguments

x	a numeric vector.
iqr_mult	interquartile range multiplier. Defaults to 3.

Details

Computes two outlier rules. The standard boxplot outlier rule based on Q1 - IQR*iqr_mult and Q3 + IQR*iqr_mult. The adjusted boxplot rule which uses the medcouple to account for data skewness. The medcouple depends on the robustbase package.

Value

• medcouple medcouple value returned from robustbase::mc(x).

• standard_rule lower and upper limits.

• adjusted_rule lower and upper limits.

References

M. Hubert; E. Vandervieren (2008). "An adjusted boxplot for skewed distributions". Computational Statistics and Data Analysis. 52 (12): 5186-5201.

See Also

mc, score_threshold

Examples

x1 = rnorm(1000,0,1)
x2 = rexp(1000,1)
boxplot_cutpoints(x=x1, iqr_mult=1.5)
boxplot_cutpoints(x=x2, iqr_mult=1.5)

ChrisDienes/SeaClass documentation built on May 18, 2019, 9:15 p.m.