GGLuca
August 6, 2024, 7:57am
1
Hello everyone,
i am having difficulties with some code that was working 6 months ago, and it fails silently now. I have tried reading the tidyverse documentation, but i am still not figuring it out.
df <- df %>%
mutate(across(everything(), ~ fct_recode(.x,
"Strongly Agree" = "5",
"Agree" = "4",
"Neither" = "3",
"Disagree" = "2",
"Strongly Disagree" = "1"
)))
where is the mistake? I mean, some suggested that across is deprecated but the documentation is still live, i wouldn't know where to begin.
Thanks!
theres no mistake unless the contents of your df are inappropriate; like having non factor columns that you then attempt to recode. also if it fails it would fail loudly unless you have altered how your R session responds to warnings and errors...
heres a small df, that shows a case where it would work
df <- data.frame(a=factor(1:5),
b=factor(5:1))
results in
a b
1 Strongly Disagree Strongly Agree
2 Disagree Agree
3 Neither Neither
4 Agree Disagree
5 Strongly Agree Strongly Disagree
>
1 Like
GGLuca
August 6, 2024, 11:29am
3
Thank you for the answer, it works. And I also think i found a culprit, namely package plyr that loaded after tidyverse creating some weird behavior. I think this thread can be closed or deleted.
Again, thank you!
My guess is that you may have been trying to aplpy fct_recode() to a numeric column rather than experiencing any issues with plyr:
library(tidyverse)
# using character column
df <- tibble(x = 1:5 |> as.character())
df %>%
mutate(across(everything(), ~ fct_recode(.x,
"Strongly Agree" = "5",
"Agree" = "4",
"Neither" = "3",
"Disagree" = "2",
"Strongly Disagree" = "1"
)))
#> # A tibble: 5 × 1
#> x
#> <fct>
#> 1 Strongly Disagree
#> 2 Disagree
#> 3 Neither
#> 4 Agree
#> 5 Strongly Agree
# using factor column
df <- tibble(x = 1:5 |> factor())
df %>%
mutate(across(everything(), ~ fct_recode(.x,
"Strongly Agree" = "5",
"Agree" = "4",
"Neither" = "3",
"Disagree" = "2",
"Strongly Disagree" = "1"
)))
#> # A tibble: 5 × 1
#> x
#> <fct>
#> 1 Strongly Disagree
#> 2 Disagree
#> 3 Neither
#> 4 Agree
#> 5 Strongly Agree
# using numeric column
df <- tibble(x = 1:5)
df %>%
mutate(across(everything(), ~ fct_recode(.x,
"Strongly Agree" = "5",
"Agree" = "4",
"Neither" = "3",
"Disagree" = "2",
"Strongly Disagree" = "1"
)))
#> Error in `mutate()`:
#> ℹ In argument: `across(...)`.
#> Caused by error in `across()`:
#> ! Can't compute column `x`.
#> Caused by error in `fct_recode()`:
#> ! `.f` must be a factor or character vector, not an integer vector.
#> Backtrace:
#> ▆
#> 1. ├─df %>% ...
#> 2. ├─dplyr::mutate(...)
#> 3. ├─dplyr:::mutate.data.frame(...)
#> 4. │ └─dplyr:::mutate_cols(.data, dplyr_quosures(...), by)
#> 5. │ ├─base::withCallingHandlers(...)
#> 6. │ └─dplyr:::mutate_col(dots[[i]], data, mask, new_columns)
#> 7. │ ├─base::withCallingHandlers(...)
#> 8. │ └─mask$eval_all_mutate(quo)
#> 9. │ └─dplyr (local) eval()
#> 10. └─forcats::fct_recode(...)
#> 11. └─forcats:::check_factor(.f)
#> 12. └─cli::cli_abort(...)
#> 13. └─rlang::abort(...)
Created on 2024-08-06 with reprex v2.0.2
I would think the plyr explanation works; I would expect it potentially overwrite dplyr::mutate with plyr::mutate, and therefore not understand accross etc...
Could be, but I was going on was:
and
but you're right — without the actual error messages, we're left to guess
system
August 17, 2024, 5:32am
8
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.