CODEWITHSUNDEEP
r
an_ja
an_ja

Reputation: 427

How to hand over a list of vectors to a function?

I have a dataframe and some of the vectors have NAs. I'd like to write a function that recodes NA to 0. However, I don't wan't to recode all vectors. I'd rather prefer specifying a list of vectors to which I can refer to when calling the function. The vectors should be recoded within the dataframe.

Here is my dataframe:

structure(list(ahvn13_chiff = structure(1:6, .Label = c("a", "b", "c", "d", "e", "f", "g", "h","i"
), class = c("labelled", "factor"), label = c(ahvn13_chiff = "Sozialversicherungsnummer anonymisiert")), 
    sex = structure(c(1L, 1L, 1L, 1L, 2L, 2L), label = c(sex = "Geschlecht"), class = c("labelled", 
    "integer")), yearmonthday_dateofbirth = structure(c(11086070400, 
    13038710400, 11838355200, 13972348800, 13959993600, 10880870400
    ), label = c(yearMonthDay_dateOfBirth = "Geburtsdatum"), class = c("labelled", 
    "numeric")), z100 = c(0L, 1331L, NA, NA, 0L, NA), z102 = c(0L, 
    222004L, 0L, 25004L, 0L, NA), z104 = c(0L, NA, 0L, NA, 0L, 
    NA), z106 = c(0L, 3604L, 0L, 0L, 0L, NA)), codepage = 65001L, row.names = c(NA, 
6L), class = "data.frame")

I managed to write a code that recodes the NAs of some of the vectors:

data[c("z100","z102")][is.na(data[c("z100","z102")])] <- 0

I then tried to convert this code to a function and ended up with these two options:

na.zero <- function(x) {
 x[is.na(x)] <- 0
}

na.zero(data[c("z100","z102")])

na.zero <- function(df,var) {
 df[var][is.na(df[var])] <- 0
}

na.zero(data, c("z100","z102"))

However, none of them works. What is wrong and how else can I hand over a list of vectors to a function.

Thank you in advance for your Inputs.

Upvotes: 0

Views: 55

Answers (1)

SebSta
SebSta

Reputation: 496

2 things:

First you are not returning anything from your functions. Add a return(x) or return(df) at the end of both functions.

Second, you would also have to assign the data again to their correct place: (Only applies to your second function)

data[c("z100","z102")] <- na.zero(data[c("z100","z102")])

Because R functions dont change an object in place.

Upvotes: 2

Related Questions