subset.survey.design: Subset of survey

Description Usage Arguments Value See Also Examples

View source: R/survey.R

Description

Restrict a survey design to a subpopulation, keeping the original design information about number of clusters, strata. If the design has no post-stratification or calibration data the subset will use proportionately less memory.

Usage

1
2
3
4
## S3 method for class 'survey.design'
subset(x, subset, ...)
## S3 method for class 'svyrep.design'
subset(x, subset, ...)

Arguments

x

A survey design object

subset

An expression specifying the subpopulation

...

Arguments not used by this method

Value

A new survey design object

See Also

svydesign

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
data(fpc)
dfpc<-svydesign(id=~psuid,strat=~stratid,weight=~weight,data=fpc,nest=TRUE)
dsub<-subset(dfpc,x>4)
summary(dsub)
svymean(~x,design=dsub)

## These should give the same domain estimates and standard errors
svyby(~x,~I(x>4),design=dfpc, svymean)
summary(svyglm(x~I(x>4)+0,design=dfpc))

data(api)
dclus1<-svydesign(id=~dnum, weights=~pw, data=apiclus1, fpc=~fpc)
rclus1<-as.svrepdesign(dclus1)
svymean(~enroll, subset(dclus1, sch.wide=="Yes" & comp.imp=="Yes"))
svymean(~enroll, subset(rclus1, sch.wide=="Yes" & comp.imp=="Yes"))

survey documentation built on July 19, 2021, 9:06 a.m.