Changing the Values with Specific Character from Data Frame in R-CodePudding

My data frame like this.

    X           Y              Z 
    10.5 m³/s   15. m³/s    14.3 m³/s
    10. m³/s    11. 5m³/s    16.7 m³/s
    10.8 m³/s   15.7 m³/s    1.5 m³/s

I have to delete some specific characters from my data frame. In order to overwhelm this issue i have wrote some code like this.

  df[] <- lapply(df,gsub, pattern='. m³/s', replacement='')

It seems to be worked. But i encountered with "." issue which is orginally being shape of the data.

X      Y       Z 
10.5   15.     14.3
10.    11.5    16.7
10.8   15.7    1.5

As well as I did every combination for the pattern function.

I need double values so i have to delete points just where placing at end of the values.

Thanks for helps.

CodePudding user response：

You can account for the decimal in the whole number values by adding a regex \\.?, which matches literal "." if present one or zero times.

d <- data.frame(X = "10.5 m³/s", Y = "15. m³/s", Z = "14.3 m³/s")
lapply(d, gsub, pattern='\\.? m³/s', replacement='')

$X
[1] "10.5"

$Y
[1] "15"

$Z
[1] "14.3"

CodePudding user response：

Maybe readr::parse_number() is also an option?

library(dplyr)
df %>% mutate(across(everything(), readr::parse_number))

Result:

# A tibble: 3 x 3
      X     Y     Z
  <dbl> <dbl> <dbl>
1  10.5  15    14.3
2  10    11    16.7
3  10.8  15.7   1.5

CodePudding user response：

library(tidyverse)

df %>% mutate(across(X:Z,~as.double(str_remove_all(.x, " |m³/s"))))

Output:

# A tibble: 3 × 3
      X     Y     Z
  <dbl> <dbl> <dbl>
1  10.5  15    14.3
2  10    11.5  16.7
3  10.8  15.7   1.5

Input:

structure(list(X = c("10.5 m³/s", "10. m³/s", "10.8 m³/s"), 
    Y = c("15. m³/s", "11. 5m³/s", "15.7 m³/s"), Z = c("14.3 m³/s", 
    "16.7 m³/s", "1.5 m³/s")), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -3L))