What's the concept behind the name and order of the argument mode?

[maurolepore]

@lindadelacombaz and @AnneSchoenauer,

This seems to be an important domain-specific concept that I would like to use consistently everywhere.

I see tiltPlot uses the name mode and the values are ordered consistently as this in this example:

plot_sankey(
  data,
  with_company = TRUE,
  benchmark = benchmarks(),
  mode = c("equal_weight", "worst_case", "best_case")
)

Can you confirm this is the best name and order?

One problem with mode is that it clashes with base::mode:

# This name is already taken in base R
mode
#> function (x) 
#> {
#>     if (is.expression(x)) 
#>         return("expression")
#>     if (is.call(x)) 
#>         return(switch(deparse(x[[1L]])[1L], `(` = "(", "call"))
#>     if (is.name(x)) 
#>         "name"
#>     else switch(tx <- typeof(x), double = , integer = "numeric", 
#>         closure = , builtin = , special = "function", tx)
#> }
#> <bytecode: 0x561357b95e58>
#> <environment: namespace:base>

So I'm considering making the name more specific by adding the prefix "weight" and making it plural, e.g.:

  weight_modes <- c("equal_weight", "worst_case", "best_case")

The plural alone solves the clash ...

# The error is a good thing. It means the name is available
modes
#> Error in eval(expr, envir, enclos): object 'modes' not found

... but still the name "modes" feel too vague

What do you think?