weighted.var: Weighted univariate variance coping with missing values

In modi: Multivariate outlier detection and imputation for incomplete survey data

Description

This function is analogue to weighted.mean.

Usage

weighted.var(x, w, na.rm = FALSE)

Arguments

x	a vector with data
w	positive weights (may not have missings where x is observed)
na.rm	if TRUE remove missing values

Details

The weights w are standardised such that ∑_{observed} w_i equals the number of observed values in x. The function calculates

∑_{observed} w_i (x_i-weighted.mean(x,w,na.rm=T))^2/((∑_{observed} w_i)-1).

Value

The weighted variance of x with weights w (with missing values removed when na.rm=TRUE).

Author(s)

Beat Hulliger

See Also

See Also as weighted.mean

Examples

x<-rnorm(100)
x[sample(1:100,20)]<-NA
w<-rchisq(100,2)
weighted.var(x,w,na.rm=TRUE)

modi documentation built on May 2, 2019, 6:48 p.m.