weightedVarRcppN: Estimate weighted variance

In carSurv: Correlation-Adjusted Regression Survival (CARS) Scores

Description

Efficient C implementation of the sample variance estimator. The denominator is defined as the sample size.

Usage

weightedVarRcppN(y, w)

Arguments

y	Response. The mean of the response contains weights (numeric vector).
w	Weights for averaging (numeric vector).

Value

Weighted variance (numeric scalar).

Note

There are no safety checks of input arguments.

Author(s)

Thomas Welchowski

Examples

# Simulate a random vector
set.seed(3975)
x <- rnorm(100)
# Calculate variance with standard implementation
varEst <- var(x) * (100-1) / 100
# Calculate weighted variance with equal weights
equalW <- rep(1, 100)
weightVarEst <- weightedVarRcppN(y=x, w=equalW)
# Output comparison
all.equal(varEst, weightVarEst)
# Runtime comparison
library(microbenchmark)
microbenchmark(Default=var(x), New=weightedVarRcppN(y=x, w=equalW), times=25)
# -> New method is multiple times faster

carSurv documentation built on May 1, 2019, 8:44 p.m.