new rope(x, complement = TRUE)

[mattansb]

New feature to report posterior probabilities above/below the rope for a more full picture.

library(bayestestR)

x <- rnorm(4000, sd = 5)

rope(x, complement = TRUE)
#> # Proportion of samples inside the ROPE [-0.10, 0.10]:
#> 
#> Inside ROPE | Above ROPE | Below ROPE
#> -------------------------------------
#> 1.61 %      |    49.08 % |    49.32 %

• add tests

[DominiqueMakowski]

I'd make that default rather than a toggleable option

[mattansb]

Yeah? I thought of this as supplementary...

[DominiqueMakowski]

I think it's pretty useful, and it's not a breaking change of the return format since we already return a dataframe so 🤷

If we make it optional it will likely end up in the abyss of forgotten features ^^

[mattansb]

Fair enough. Then should this even be optional? Should complement = FALSE be an option at all?

[DominiqueMakowski]

I'd say no :p
Let's just document in the docstrings what are these columns and how to interpret them / how useful they can be

[strengejacke]

Have you checked if this breaks any other packages? (parameters::model_parameetrs() for Bayesian?)

[strengejacke]

Agree to include it in general and have no option.