Split plot into facets when predictions for all single terms are requested

[strengejacke]

This plot looks a bit strange, I think it makes more sense, at least to have an option to split into facets.

library(modelbased)
fit <- lm(mpg ~ cyl + disp + gear, data = mtcars)

result <- estimate_relation(fit)
plot(result)

^{Created on 2025-01-10 with reprex v2.1.1}

[DominiqueMakowski]

it does look meg, but then why would you by all 3 predictors at the same time if there's no interaction... I don't think it's something people would meaningfully use in real life

[bwiernik]

Do we want to default to just plotting the first predictor unless the user explicitly asks for 2 or 3?

[DominiqueMakowski]

plotting what is in estimate_relation() is okay I think, it's just that in this case people should estimate_relation(by="cyl") only (for instance) in the first place, as having the combination of all the factors is useless

[strengejacke]

it does look meg, but then why would you by all 3 predictors at the same time if there's no interaction... I don't think it's something people would meaningfully use in real life

Agree, but I think this was one of the examples taken from the docs or vignettes

[DominiqueMakowski]

I'll be going through the vignettes now and will do some spring cleaning :) we'll see what needs to be adjusted