variable_weight: variable_weight

In lorenc5/Rweights3: Rweights3 -- Poststratification Weighting

Description

Recode an existing variable to its “weight to” proportions

Usage

variable_weight(vector, var_weight_to)

Arguments

vector	Numeric vector from unweighted data.
var_weight_to	Numeric vector. Probably object from “weight amount” function.

Value

Numeric vector of weights applied to each respondent in the dataset.

Author(s)

Loren Collingwood <[e-mail]>

Examples

#NOT RUN: READ IN DATA
data (wapoll)

#NOT RUN: CREATE WEIGHT TO PROPORTIONS
weight_props_age <- weight_to(c(26.0,39.3,34.7), c("18-40","41-65","66+"))
weight_props_gender <- weight_to(c(49,51), c("Male","Female"))

#NOT RUN: CREATE PROPORTIONS FROM DATA
age_current <- gen_survey_props(wapoll$age3) #1=18-40,2=41-65,3=66+; age3
gender_current <- gen_survey_props(wapoll$gender) #0=Male, 1=Female

#NOT RUN: CREATE WEIGHT TO AMOUNT
age_weight_to <- weight_amount(weight_props_age,age_current)
gender_weight_to <- weight_amount(weight_props_gender,gender_current)

#NOT RUN: CREATE AGE*GENDER WEIGHT
age_weight_2 <- variable_weight(wapoll$age3,age_weight_to)	
summary(age_weight_2)
gender_weight_2 <- variable_weight(wapoll$gender,gender_weight_to)	
summary(gender_weight_2)

lorenc5/Rweights3 documentation built on May 29, 2019, 11:05 p.m.