Why is dfidx::filter() prefered over tidytable::filter()?

tidyverse
1

I hope tidytable questions belong here. My table has over a million rows and I decided to switch from dplyr/tidyr to tidytable. Unfortunately, this gives strange results when I combine filter() and n(). It worked perfectly fine using dataframes and dplyr, albeit slow.

library(tidytable)

df1 <- data.frame(a = 1:6,
                  b = c(1, 1, 1, 1, 2, 3))
td1 <- as_tidytable(df1)
td1 |>
  group_by(b) |>
  filter(n() > 1) |>
  ungroup()

This gives the following error:

Error in filter():
:information_source: In argument: n() > 1.
Caused by error in n():
! n() should only be used inside tidytable verbs

I used NCmisc::list.functions.in.file() and found that dfidx::filter() is prefered over tidytable::filter(). I managed to make it work by adding this on the second line:

conflicted::conflict_prefer_all(winner = "tidytable")

Can someone please explain to me what is going on? Why is the package dfidx prefered even though I never loaded it? This gives me an unsettling feeling that I am making other mistakes in tidytable.

2

I can't replicate this issue.

library(tidytable)
#> Warning: package 'tidytable' was built under R version 4.4.3
#> 
#> Attaching package: 'tidytable'
#> The following objects are masked from 'package:stats':
#> 
#>     dt, filter, lag
#> The following object is masked from 'package:base':
#> 
#>     %in%

df1 <- data.frame(a = 1:6,
                  b = c(1, 1, 1, 1, 2, 3))
td1 <- as_tidytable(df1)
td1 |>
  group_by(b) |>
  filter(n() > 1) |>
  ungroup()
#> # A tidytable: 4 × 2
#>       a     b
#>   <int> <dbl>
#> 1     1     1
#> 2     2     1
#> 3     3     1
#> 4     4     1

Created on 2025-04-07 with reprex v2.1.1

3

Nor can I.

Kikkervelf
Is there a change you have something in your environment that is interfering?Perhaps a conflict between two packages.

4

Thank you for your replies. I don't get the error message anymore and don't know what went wrong, but at least it's gone now.