Hi, and thanks for your great work on {tibble} and {pillar}.
In my daily workflow, I often create intermediate non-informative columns used only during processing steps (temporary variables, helper columns for joins, reshaping, etc.).
To clearly mark them as such, I prefix their names with a ..
This helps keep things tidy and indicates these variables are not meant for display or interpretation.
However, when printing a tibble, these columns clutter the output even though they are not useful for human interpretation.
I'd like to propose a feature where such columns could be hidden by default from the print output, based on a configurable prefix :
options(pillar.hide_column_prefix = ".")
This would make the output more readable without losing data.
I could override print.tbl_df() in my code, but that would be very hacky and not maintainable.
Notes
- Of course, this should be opt-in and have no impact on default behavior.
- I'm not 100% sure "." is the best prefix (maybe
.., _, or even tmp_ ?).
Ultimately, this is up to the user, but I'd be happy to get your opinion about this.
- Maybe an argument in
print() to opt-out would be nice : df |> print(n=Inf, print_all=TRUE)
Thanks again for considering this!
Hi, and thanks for your great work on {tibble} and {pillar}.
In my daily workflow, I often create intermediate non-informative columns used only during processing steps (temporary variables, helper columns for joins, reshaping, etc.).
To clearly mark them as such, I prefix their names with a
..This helps keep things tidy and indicates these variables are not meant for display or interpretation.
However, when printing a tibble, these columns clutter the output even though they are not useful for human interpretation.
I'd like to propose a feature where such columns could be hidden by default from the print output, based on a configurable prefix :
This would make the output more readable without losing data.
I could override
print.tbl_df()in my code, but that would be very hacky and not maintainable.Notes
..,_, or eventmp_?).Ultimately, this is up to the user, but I'd be happy to get your opinion about this.
print()to opt-out would be nice :df |> print(n=Inf, print_all=TRUE)Thanks again for considering this!