Error: Problem with `mutate()` input `..1`

[JojoSouza]

I don't know why, but not being able to generate the values ​​for one id/week/category/dt specifically, as you can see below, as well as not being able to generate for all data using my f2 function, it gives error in mutate.. What am I doing wrong?

Code executable

library(dplyr)

df1<-structure(list(Id = c(1, 1, 1, 1, 1, 1), date = structure(c(19090, 
19090, 19090, 19090, 19090, 19090), class = "Date"), date2 = structure(c(18962,18962, 18962, 18962, 18962, 18962), class = "Date"), 
Week = c("Wednesday","Wednesday", "Wednesday", "Wednesday", "Wednesday", 
"Wednesday"), DT = c(1, NA_character_, NA_character_,NA_character_, NA_character_, 1), Category = c("AB","CD", "EF", "GH", "IJ", "KL"), 
Time = c(1.86, 4, 2.89, 0, 1, 2.4)), row.names = c(NA, -6L), class = c("tbl_df","tbl", "data.frame"))

> df1
# A tibble: 6 x 7
     Id date       date2      Week      DT    Category  Time
  <dbl> <date>     <date>     <chr>     <chr> <chr>    <dbl>
1     1 2022-04-08 2021-12-01 Wednesday 1     AB        1.86
2     1 2022-04-08 2021-12-01 Wednesday NA    CD        4   
3     1 2022-04-08 2021-12-01 Wednesday NA    EF        2.89
4     1 2022-04-08 2021-12-01 Wednesday NA    GH        0   
5     1 2022-04-08 2021-12-01 Wednesday NA    IJ        1   
6     1 2022-04-08 2021-12-01 Wednesday 1     KL        2.4 
    

f2 <- function(df1,idd,ds,codagr,dt) {
   
  nms <- c('Time|time')
  
  mtime <- df1 %>%
    group_by(Id,Week = tools::toTitleCase(Week), Category,DT) %>% 
    summarise(across(matches(nms), mean, .names = 'Time',na.rm = TRUE), .groups = 'drop')
  
  mtime <- transform(mtime, Time = format(round(Time, digits = 2), nsmall = 2))           
  
  mtime <- mtime %>%
    filter(Id==idd,Week == ds, Category == codagr,DT==dt)
  
  return(mtime)
}

---

Generate to specific id/week/category/dt

f2(df1,"1","Wednesday","AB","1")
  Id      Week Category DT Time
  1  1 Wednesday       AB  1 1.86

f2(df1,"1","Wednesday","IJ",NA)
[1] Id       Week     Category DT       Time    
<0 linhas> (ou row.names de comprimento 0)

---

Generate to all

Result <- df1 %>%
rowwise()
mutate(f2(df1,Id,Week,Category, DT))%>%
data.frame()

Error: Problem with `filter()` input `..1`.
i Input `..1` is `Id == idd`.
x object 'Id' not found

Result correct output:

  Id      Week Category   DT Time
1  1 Wednesday       AB    1 1.86
2  1 Wednesday       CD <NA> 4.00
3  1 Wednesday       EF <NA> 2.89
4  1 Wednesday       GH <NA> 0.00
5  1 Wednesday       IJ <NA> 1.00
6  1 Wednesday       KL    1 2.40

[yifanliu]

@JojoSouza

Result <- df1 %>%
rowwise()
mutate(f2(df1,Id,Week,Category, DT))%>%
data.frame()

Error: Problem with `filter()` input `..1`.
i Input `..1` is `Id == idd`.
x object 'Id' not found

The error is due to you didn't input any .data to mutate()
And I don't think there seems to be a need to use mutate to quote the functionf2 here. directly do:

df1 %>%
  group_by(Id,Week, Category,DT) %>% 
  summarise(across(matches(nms), mean, .names = 'Time',na.rm = TRUE), .groups = 'drop') %>% 
  mutate(Time = format(round(Time, digits = 2), nsmall = 2))

# A tibble: 6 x 5
     Id Week      Category DT    Time 
  <dbl> <chr>     <chr>    <chr> <chr>
1     1 Wednesday AB       1     1.86 
2     1 Wednesday CD       NA    4.00 
3     1 Wednesday EF       NA    2.89 
4     1 Wednesday GH       NA    0.00 
5     1 Wednesday IJ       NA    1.00 
6     1 Wednesday KL       1     2.40 

is feasible in my opinion.

[JojoSouza]

Thanks for the answer, I thought the way you said it too, however, I need to somehow make it call the f2 function. Do you think it's possible?

[yifanliu]

it is possible, but requires some modifications.
Because na values are not able to do the logic operation therefore can't filter:

> NA==NA
[1] NA
> filter(df1,DT==NA)
# A tibble: 0 x 7
# ... with 7 variables: Id <dbl>, date <date>, date2 <date>, Week <chr>, DT <chr>, Category <chr>, Time <dbl>

So, I just replaced all NAs with character "NA".

f2 <- function(idd,ds,codagr,dt) {
  
  nms <- c('Time|time')
  
  mtime <- df1 %>% mutate(DT = replace_na(DT, "NA")) %>% 
    filter(Id==idd,Week == ds, Category == codagr,DT == dt) %>% 
    group_by(Id,Week,Category,DT) %>% 
    summarise(across(matches(nms), mean, .names = 'Time',na.rm = TRUE), .groups = 'keep') %>% 
    mutate(Time = format(round(Time, digits = 2), nsmall = 2))

  return(mtime)
}

df1 %>% mutate(DT = replace_na(DT, "NA")) %>% 
  rowwise %>% mutate(f2(idd=Id,ds=Week,codagr=Category, dt=DT),.keep = 'none')

# A tibble: 6 x 5
# Rowwise: 
     Id Week      DT    Category Time 
  <dbl> <chr>     <chr> <chr>    <chr>
1     1 Wednesday 1     AB       1.86 
2     1 Wednesday NA    CD       4.00 
3     1 Wednesday NA    EF       2.89 
4     1 Wednesday NA    GH       0.00 
5     1 Wednesday NA    IJ       1.00 
6     1 Wednesday 1     KL       2.40 

do filter in f2 before summarise may saves time.

[JojoSouza]

Thanks for reply @yifanliu