Transforming some codes on weighted chi-square tests into functions of variables-CodePudding

I have a code that does weighted chi-square tests. It also takes the p-value and add it into a dataset. I would like to transform the code into a function where the arguments is the variable names. The data (called) looks something like that:

Gender  UserStatus  Weight
Female  HC          1.3
Male    IDP         0.8
Male    HC          1.3
Male    V           1.1
Female  IDP         0.8
Female  HC          1.3
Female  IDP         0.8
Male    HC          1.3
Female  IDP         0.8
Male    IDP         0.8
Male    IDP         0.8
Female  IDP         0.8
Female  V           1.1
Male    IDP         0.8
Male    V           1.1
Female  HC          1.3
Male    IDP         0.8
Male    V           1.1
  ....

The code without the function looks like this:

#Create the empty database
dTests <- data.frame(matrix(ncol = 3, nrow = 0))
colnames(dTests) <- c('var1', 'var2', 'pvalue')

#Affiliate the weight to the data
dGenderUserStatus <- data %>%
  filter(!is.na(Gender) & !is.na(UserStatus)) %>%
  as_survey_design(weights = Weight)

#Calculate the Chi square p-value
pvalue <- svychisq( ~ Gender   UserStatus, dGenderUserStatus) %>% 
.$p.value

#add row
dTests[nrow(dTests)   1,] <- c("Gender","UserStatus", pvalue)

The function - that I attempted to write and for which it does not work - looks something like that:

   #Create the empty database
    dTests <- data.frame(matrix(ncol = 3, nrow = 0))
    colnames(dTests) <- c('var1', 'var2', 'pvalue')
    
FctChiSquareTestGender <- function(data,Var,weight){
  dGenderVar <- data %>%
    filter(!is.na(Gender) & !is.na(get(Var))) %>%
    as_survey_design(weights = {{weight}})

  pvalue <- svychisq( ~ Gender   get(Var), dGenderVar) %>% 
    .$p.value

  #add row
  dTests[nrow(dTests)   1,] <- c("Gender",Var, pvalue)

    }
FctChiSquareTestGender(data,"UserStatus",Weight)

Unfortunately, the code does not work from the pvalue creation with the function "svychisq" when I call "get(Var)". I get this error message:

Do you have any idea on what I should do about get(Var) so that this code works?

Thanks in advance,

Stephanie

CodePudding user response：

I suspect you don't need to use get(Var) in the formula, you can just use the string Var:

Gender <- rbinom(100, 1, 0.5)
statuses <- c("HC", "IDP", "V")
Userstatus <- replicate(sample(statuses, 1), n = 100)

myfunction <- function(Var){
  form <- as.formula(paste("~ Gender", Var, sep = " "))
  form
}
myfunction("Userstatus")
#> ~Gender   Userstatus
#> <environment: 0x55ba8badbbf0>

And you put form in place of the formula input to svychisq().