R: Round a data frame with a specific formula across all columns with dplyr-CodePudding

I would like to use the following formula to round my whole data frame in R:

format(round(Df$`p-value`, digits=2), nsmall = 2)

Now, the code above is only for one column (just to show you).

I know dplyr is easy to use when I want to use the formula for my whole data frame as there is the across() function. I tried the following:

Df %>%
mutate(across(format(round(digits=2), nsmall = 2)))

However, I get an error, saying that "argument "x" is missing, with no default". I assume I need to make a small adjustment to the code but I couldn't figure out what I need to adjust? Could someone help me here please?

Thank you.

CodePudding user response：

This should do it. You need to specify the .fns argument by name or the formula will be matched by position to the first argument of across, .cols.

DF %>% mutate( across( .fns = ~ format(round(.x, digits = 2), nsmall = 2) )

CodePudding user response：

You forgot argument x in it. R is completely right. You need to get the dataframe (argument x) in the function which you want to convert. In this case a "." represents the dataframe, because its dplyr. You mutate(across) was not working for me, so I decided to use mutate_all, which does the job perfectly.

A %>%    mutate_all(~format(round(., digits=1), nsmall = 2))

CodePudding user response：

First I create some random data:

Df <- data.frame(v1 = c(1.21,3.21,2.2,1.11),
                 v2 = c(1.333, 2.22,1.1111,2.1))

You can use the following code:

library(dplyr)
Df %>%
  mutate(across(everything(~format(round(digits=2), nsmall = 2))))

Output:

    v1     v2
1 1.21 1.3330
2 3.21 2.2200
3 2.20 1.1111
4 1.11 2.1000