srs.domainestimator {samplingR}

Simple Random Sample parameter estimation of domains.

Description

Function to make estimations of diferent parameters on a given domain based on a Simple Random Sample.

Usage

srs.domainestimator(
  Nh,
  data,
  estimator = c("total", "mean", "proportion", "class total"),
  domain,
  replace = FALSE,
  alpha
)

Arguments

Nh

Number of instances of the data set domain.

data

Sample of the data. It must constain a column with the data to estimate and a second column with the domain of each instance.

estimator

One of "total", "mean". Default is "total".

domain

Domain of the sample from which parameter estimation will be done.

replace

Whether the sample to be taken can have repeated instances or not.

alpha

Optional value to calculate estimation error and build 1-alpha confidence interval.

Details

Data columns must be arranged with interest values on the first column and domain values on the last column.
Domain parameter can be either numeric or character and must be equal to one of the values of the domain column of data.

Value

A list containing different interest values:

• estimator

• variance

• sampling.error

• estimation.error

• confint

References

Pérez, C. (1999) Técnicas de muestreo estadístico. Teoría, práctica y aplicaciones informáticas. 193-195

Examples

data<-cbind(rnorm(500, 50, 20), rep(c(1:2),250))
sample<-data[srs.sample(500, 100),]
sum(data[which(data[,-1]==1),1])
srs.domainestimator(Nh = 250, data = sample, estimator="total", domain=1)