0

I'm trying to create a function to report number of missing values of one variable. My code is as below:

var1=c(1:10,rep(NA,5)) var2=c(rep(NA,10), 21:25) var3=as.factor(c(rep(NA,3), rep("V",8), rep("U",4))) var4<-c(rep(NA,3), 1:9,rep(NA,3)) data <- data.frame(var1,var2,var3,var4)

So, the codes below gave me what I want, which is the count of missing values in one variable given the other variable's not missing: data %>% filter(!is.na(var3)) %>% summarise(Missing=sum(is.na(var2))) data %>% filter(!is.na(var2)) %>% summarise(Missing=sum(is.na(var4)))

But when I put them in a function, it didn't work:

count_missing <- function(data,a,b) {
  data %>% filter(!is.na(a)) %>% summarise(Missing=sum(is.na(b)))
}

For example: count_missing(data,var2,var4) should give me:

Missing
1       3

But instead, it returned:

Missing
1       6

My guess is the summarise(Missing=sum(is.na(b)) in the function doesn't take input from the pipeline.

Can anyone help me with this? Thank you!

Vu Tran
  • 1

1 Answers1

0

Rewrite your function using the {{}} operator:

count_missing <- function(data,a,b) {
  data %>%
    filter(!is.na({{a}})) %>%
    summarise(Missing=sum(is.na({{b}}))) 
}

Usage:

count_missing(data,var2,var4)

Output:

  Missing
1       3
langtang
  • 22,248
  • 1
  • 12
  • 27
  • Hi langtang, Thank you for your prompt response! I've tried your suggestion and it threw an error Error in UseMethod("filter") : no applicable method for 'filter' applied to an object of class "c('integer', 'numeric')" . – Vu Tran Jun 02 '22 at 02:21
  • Do I have to install any package to use {{}}? I have rlang 0.4.11 already. Thank you! – Vu Tran Jun 02 '22 at 02:28
  • I've tried it again and it worked. Thank you very much! – Vu Tran Jun 02 '22 at 02:50