[Feature Request]: Add a new format

[mivanov5]

Feature description

In the list of possible formats for ratables/tern functions (produced by list_valid_format_labels(), formatters package) the following format is missing:
(xx. - xx.)
We have similar formats like (xx., xx.) and xx. (xx.%), but not the one above.

Please add this format.

Code of Conduct

• I agree to follow this project's Code of Conduct.

Contribution Guidelines

• I agree to follow this project's Contribution Guidelines.

Security Policy

• I agree to follow this project's Security Policy.

[clarkliming]

I do have a idea that we somehow redesign the "xx.x" things, to make it flexible (and translate it into "sprintf"). e.g.,

"sadbasd asa a xx.xx asda a" translate to "sadbasd asa a %.2f asda a"

[Melkiades]

This possibility exists and it is working very fine:

library(rtables)
fmtfun <- sprintf_format("this range => %.1f")
format_value(100, format = fmtfun)

fmtfun2 <- sprintf_format("sadbasd asa a %.2f asda a %.1f")
format_value(list(12.23456, 2.724), format = fmtfun2)

lyt <- basic_table() %>%
  split_cols_by("Species") %>%
  analyze("Sepal.Length", afun = function(x) {
    list(
      "mean (sd)" = rcell(c(mean(x), sd(x)), format = fmtfun2),
      "range" = rcell(diff(range(x)), format = fmtfun)
    )
  })

> build_table(lyt, iris)
                       setosa                        versicolor                       virginica          
—————————————————————————————————————————————————————————————————————————————————————————————————————————
mean (sd)   sadbasd asa a 5.01 asda a 0.4   sadbasd asa a 5.94 asda a 0.5   sadbasd asa a 6.59 asda a 0.6
range             this range => 1.5               this range => 2.1               this range => 3.0      

I think this MUST be better documented though as it is a very powerful and elegant solution to multiple little problems

[clarkliming]

oh I mean we have two sets of printing format, one is from "printf_format", one is translating "xx.xx" into "%.2f" so that users could use any "xx.xx" to construct there format (as this is more common in practice for tlg works)

[Melkiades]

I see what you mean. I think it is a bit of overhead for little gain as you would use directly sprintf_format if you need to do that. I think the way format currently handles "xx" is a bit more complete than its printf cousin as it handles strings and NAs for any rcell specifically. I think doing what you suggested will end up losing functionality. I could be wrong though, maybe @gmbecker thinks differently

[gmbecker]

This essentially falls under factoring formatters format detection into a parser, which is on the roadmap, but a large piece of work we have not yet had the resources to pursue.

[shajoezhu]

hi @mivanov5 , I think we already have the work-around in #880 (comment), would this work for you?

We can look into refactoring and improvement for future work. Thanks!

[mivanov5]

Hi @shajoezhu, I made another modification to my program, but guess that as suggested above next time I can create my own format and use its name instead of "xx. - xx.". Hope that anyway it will be added to the formats list at some point. Thanks to @Melkiades for the example and to everyone else for your comments.