[R] Calibrate survey using rake

Juan Ceccarelli Arias j|c@283 @end|ng |rom gm@||@com
Tue Jul 9 17:19:29 CEST 2019


Hello,
I need to calibrate a survey using raking.
I already declared the complex design here:

des_survey <- svydesign(id= ~idd, strata= ~strata, data = datos, weights =
~ft, check.strata=TRUE  )
options(survey.lonely.psu = "remove")
options(digits = 10)

However, when I run the rake process, the tabulates using the original
weights and the calibrated doesn't match to the population marginals I
declared.

des_surveyR <- rake(diseño, ~region, ~pop.region)

The population marginals is only one. Region is the variable. The data
frame pop.region containes the total to calibrate.

Can you guide me in order to calibrate?

	[[alternative HTML version deleted]]



More information about the R-help mailing list