Create variables using mutate within a loop

Question

I have a data frame with x columns. From each column, I want to create a new column, where data are replaced by their ranks, but the ranks are not standard but follow a pyramid style.

To do this, I am tryng to use mutate within a function and loop that function over the different columns.

I have tried this in different ways. However, I keep on getting an error (see below code) and I do not understand why I get that error.

df <- tibble(Z1 = rnorm(40), Z2 = rnorm(40))
pyramid = c(rep(5,2),rep(4,3),rep(3,3), rep(2,4), rep(1,5), rep(0,6), 
                    rep(-1, 5), rep(-2,4), rep(-3, 3), rep(-4,3), rep(-5,2))


for (i in 1:2){
  Z <- rlang::sym(paste("Z", i, sep=""))
  QS <- rlang::sym(paste("QS", i, sep=""))
  
  df <- df %>% arrange(!!-Z) %>% mutate(!!QS = pyramid)
}

The error:

Error: unexpected '=' in:
"  
  df <- df %>% arrange(!!-Z) %>% mutate(!!QS ="
> }
Error: unexpected '}' in "}"

Since the following code is working, I suspect it comes from the way I use a symbol to create a new variable. I have also tried using a String and got the same error. I have also tried the double curly instead of the double !! and it got the same issue So I am lost!

for (i in 1:2){
  Z <- rlang::sym(paste("Z", i, sep=""))

  df <- df %>% arrange(!!-Z) %>% mutate(QS2 = pyramid)
}

The expected output should look like this:

# A tibble: 40 x 4
       Z1    Z2   QS1   QS2
    <dbl> <dbl> <dbl> <dbl>
 1 -0.591 1.64     -2     5
 2 -0.132 1.59      0     5
 3 -0.418 1.59     -2     4
 4  1.11  1.52      4     4
 5  1.65  1.15      4     4
 6  0.289 1.11      1     3
 7  1.85  1.09      5     3
 8  0.526 1.07      1     3
 9 -0.436 1.04     -2     2
10 -0.671 0.794    -3     2
# ... with 30 more rows

Sorry, I should have done so. I am editing the question to show the expected results; In short I am transforming each value into an index. This index is not linear, because you will have two 5, three 4, etc... This is because I am analysing the results of ranking made by respondents that were made this way (Q-methodologyà — djourd1
– djourd1, Commented Jun 28, 2020 at 7:03
If you use unquoting on the LHS of an assignment you have to replace ´=` by :=, i.e. mutate(!!QS := pyramid). — stefan
– stefan, Commented Jun 28, 2020 at 7:28

Ronak Shah · Accepted Answer · 2020-06-28 07:16:26Z

1

You can use order in mutate :

library(dplyr)
df %>% mutate(across(everything(), list(QS = ~pyramid[order(.)])))
#In old dplyr
#df %>% mutate_all(list(QS = ~pyramid[order(.)]))

Or in base R :

df[paste0('QS', seq_along(df))] <- lapply(df, function(x) pyramid[order(x)])

answered Jun 28, 2020 at 7:16

Ronak Shah

391k20 gold badges173 silver badges237 bronze badges

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

1 Comment

djourd1 Over a year ago

Indeed, much more compact !

akrun · Accepted Answer · 2020-06-28 19:26:34Z

1

An option with data.table

library(data.table)
setDT(df)[, paste0('QS', names(df)) := lapply(.SD, function(x) pyramid[order(x)])]

answered Jun 28, 2020 at 19:26

akrun

891k38 gold badges590 silver badges700 bronze badges

Collectives™ on Stack Overflow

Create variables using mutate within a loop

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related