Hb: Detection of outliers using the Hidiroglou-Berthelot (HB)...

In statisticsnorway/Kostra: Functions for Kostra

Detection of outliers using the Hidiroglou-Berthelot (HB) method

Description

Detects possible outliers of a variable in period t by comparing it with revised values from period t-1

Usage

Hb(data, id, x1, x2, pU = 0.5, pA = 0.05, pC = 20, strataName = NULL)

Arguments

data	Input of Hb is a data set of class data.frame.
id	Name of an identification variable.
x1	Name of variable in period t.
x2	Name of variable in period t-1.
pU	Parameter that adjusts for different level of the variables.Default value 0.5.
pA	Parameter that adjusts for small differences between the median and the 1st or 3rd quartile. Default value 0.05.
pC	Parameter that controls the length of the confidence interval.Default value 20.
strataName	Name of stratification variable. Optional. If strataName is given, the HB method is performed within each stratum.

Value

Output of Hb is a data set of class data.frame. All units are returned, but the HB method is only performed on the data set where units with both x1 and x2 not missing and greater than zero are included. In this data set, units with x1 = x2 are included in the HB method only if they cover less than 50 per cent of the number of units in the stratum. The output variables are:

id	The input identification variable
x1	The input x1 variable
x2	The input x2 variable
maxX	The maximum of x1 and x2
ratio	The ratio between x1 and x2
lowerLimit	The lower limit of the ratio
upperLimit	The upper limit of the ratio
medRatio	The median ratio
outlier	A binary variable indicating whether the observation is an outlier (1) or not (0)
strata	Strata name or number

Author(s)

Magnar Lillegård

References

Datarevisjon: Kontroll, granskning og retting av data. Anbefalt praksis, Statistisk sentralbyrås håndbøker, 2005/84.

Examples

 minedata <- KostraData("testdata")
 minedata$strata <- as.character(c(rep(1, 61), rep(2, 91), rep(3, 98), rep(4, 81), rep(5, 85)))
 Hb(data = minedata, id = "Region", x1 = "areal_381_eier_2015", x2 = "areal_381_eier_2014")
 Hb(data = minedata, id = "Region", x1 = "areal_381_eier_2015", x2 = "areal_381_eier_2014",
    strataName = "strata")
 Hb(data = minedata, id = "Region", x1 = "areal_381_eier_2015", x2 = "areal_381_eier_2014",
    pA = 0.1, strataName = "strata")

statisticsnorway/Kostra documentation built on Nov. 2, 2024, 6:40 p.m.