weighted.quantile function

Quantiles of a weighted cdf

A weighted cdf is calculated and quantiles are evaluated. Missing values are discarded.

weighted.quantile(x, w, prob = 0.5, plot = FALSE)

Arguments

• x: a vector of data.
• w: a vector of (sampling) weights.
• prob: the probability for the quantile.
• plot: if TRUE, the weighted cdf is plotted.

Returns

The quantile according to prob (by default it returns the weighted median).

Details

Weighted linear interpolation in case of non-unique inverse. Gives a warning when the contribution of the weight of the smallest observation to the total weight is larger than prob.

Note

No variance calculation.

Examples

x <- rnorm(100)
x[sample(1:100, 20)] <- NA
w <- rchisq(100, 2)
weighted.quantile(x, w, 0.2, TRUE)

See Also

svyquantile

Author(s)

Beat Hulliger