The Horvitz-Thompson estimator and the Hartley-Rao approximation require information in addition to the sampling probabilities for sampled individuals. These functions allow this information to be supplied.

HR(psum=NULL, strata = NULL)
ppsmat(jointprob, tolerance = 1e-04)
ppscov(probcov, weighted=FALSE)

Arguments

psum

The sum of squared sampling probabilities for the population, divided by the sample size, as a single number or as a vector for stratified sampling

strata

Stratum labels, of the same length as psum, if psum is a vector

jointprob

Matrix of pairwise sampling probabilities for the sampled individuals

tolerance

Tolerance for deciding that the covariance of sampling indicators is zero

probcov

Covariance of the sampling indicators (often written 'Delta'), or weighted covariance if weighted=TRUE

weighted

If TRUE, the probcov argument is the covariance divided by pairwise sampling probabilities

Value

An object of class HR,ppsmat, ppsdelta, or ppsdcheck suitable for supplying as the pps argument to svydesign.

See also

election for examples of PPS designs

Examples

HR(0.1)
#> $pbar
#> $pbar[[1]]
#> [1] 0.1
#> 
#> 
#> $strata
#> $strata[[1]]
#> [1] 1
#> 
#> 
#> $call
#> HR(0.1)
#> 
#> attr(,"class")
#> [1] "HR"