conditionally replacing values of a series of variables using a loop [duplicate]

Question

I am converting extensive stata code to R.

In stata, if I have a series of variables such as var1, var2, var3, etc., and I want to change (recode) a specific value if it appears in any of the variables, I can do that using the statement recode var*(9999 = -9). In this case, I want to change all 9999's in the variable series to NA.

HERE IS THE CODE I TRIED

data = data.frame(var1=c(5, 56, 9999, 56, 78, 51),
              var2=c( 9999, 56, 43, 56, 78, 9999),
              var3=c(5, 34, 56, 78, 76, 79))
 varlist=gsub(" ","",paste("data$var",1:3,sep=""))
 varlist
 summary(data$var2)
 for (v in varlist){
   v[v=="9999"] = NA
 }
summary(data$var2)
data$var2[data$var2==9999] = NA
summary(data$var2)

LOOP DIDN'T WORK, SINGLE ASSIGNMENT (data$var2[data$var2==9999] = NA) DID AS SHOWN BY WHAT THE THREE summary(data$var2) SHOW:

Min. 1st Qu. Median Mean 3rd Qu. Max. 43 56 67 3372 7519 9999

Min. 1st Qu. Median Mean 3rd Qu. Max. NA's 43.00 52.75 56.00 58.25 61.50 78.00 2

THE CONTENTS OF VARLIST IS [1] "data$var1" "data$var2" "data$var3"

I ALSO TRIED THE FOLLOWING LOOP BASED ON WHAT I FOUND ON STACKOVERFLOW (https://forum.posit.co/t/using-variables-names-in-loops/128653/2):

for (i in 1:3) {
  variable = paste0("data$var", i)
  variable[variable==9999] = NA
}

IT ALSO DIDN'T WORK.

CLEARLY I'M MISSING SOMETHING HERE.

Welcome to SO, but you don't make friends by shouting. In the tidyverse, somethings simliar to mutate(across(starts_with("var"), \(x) ifelse(x == 9999, NA, x)) should give you what you want. — Limey
– Limey, Commented Aug 7, 2024 at 21:09
Wasn;'t shouting ... when I wrote the message, I originally tried to distinguish code by using lower case not realizing it would be reformatted ... — Bert Kritzer
– Bert Kritzer, Commented Aug 8, 2024 at 14:26
Thank you, that worked!! Still need to understand what I can and cannot in terms of column/variable manipulation using a loop; can you direct me to a relevant source? — Bert Kritzer
– Bert Kritzer, Commented Aug 8, 2024 at 14:35

Fabio Correa · Accepted Answer · 2024-08-07 21:10:27Z

1

Try: data[data==9999] <- NA

answered Aug 7, 2024 at 21:10

Fabio Correa

1,4511 gold badge13 silver badges21 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

conditionally replacing values of a series of variables using a loop [duplicate]

1 Answer 1

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Linked

Related