Counting NA's of variables with for loop-CodePudding

I would like to count the NA of a set of variables with a for loop.

confounders<-c("Age", "Sex",  "Race", "education")

missings = list() 
for (var in confounders) { missings[[var]] = with(data_all, sum(is.na(var[[1]])))} 
missings

I only get the output zero for every variable, which is incorrect:

> missings
$Age
[1] 0

$Sex
[1] 0

$Race
[1] 0

$education
[1] 0


> with(data_all, sum(is.na(education)))
[1] 55

CodePudding user response：

You need to tell R to evaluate var a variable, not as a string (e.g. Age, not "Age"). You can do this with get(eval())

data_all=data.frame(Age=c(34,40,50,40),Sex=c("F","M","M","F"), Race=c(NA,NA,NA,NA), education=c("high","low","low",NA))

confounders<-c("Age", "Sex","Race", "education")

missings = list() 
for (var in confounders) { 
    missings[[var]] = with(data_all, sum(is.na(get(eval(var)))))
} 
missings

CodePudding user response：

Try this

confounders<-c("Age", "Sex",  "Race", "education")

missings = list() 
for (var in confounders) { 
             missings[[var]] = sum(is.na(data_all[[var]])) 
             } 
missings