Remove redundant x/y-axis in facets

[strengejacke]

Not sure if this is beyond the package's scope, but it would be great if tinyplot could remove redundant x- and y-axes in facets. See example below, where we only need the "outer" axes. Or can this (easily?) be done manually with base R graphics tweaks?

library(modelbased)
library(tinyplot)
data(efc, package = "modelbased")
m <- lm(neg_c_7 ~ c172code * e42dep, data = efc)

op <- tpar(
  bty = "n", # No box (frame) around the plot
  grid = TRUE, # Add a background grid
  las = 1, # Horizontal axis tick labels
  pch = 16 # Filled points as default
)

estimate_means(m, c("c172code", "e42dep")) |> plt(facet = ~e42dep, dodge = 0)

^{Created on 2026-07-06 with reprex v2.1.1}

[grantmcdermott]

Hey @strengejacke, it's a bit hidden in the documentation, but just include tinyplot(..., frame = FALSE) among your arguments.

I've been meaning to add a dedicated argument that could get passed to facet.args, but didn't have time before the 0.7.0 release. Let's keep this open as a reminder for when I get back from vacation.

P.S. You might like the new "float" theme, which gives the look you want plus some added extras. Edit: See here for an example of a tweaked float2 theme (with an added background grid). https://grantmcdermott.com/tinyplot/man/tinytheme_register.html#examples

[zeileis]

This could also be something for the Gallery and/or the Tips & Tricks vignette.

Supporting the functionality in type_spineplot() would also be good, will add it to #660

[strengejacke]

Great, thanks! I'll just add frame = FALSE by default to tinyplots from modelbased.

A short follow-up question, a bit off-topic: for few categories, like in my above example, is it possible to have slightly larger margins/paddings at the left end right end of the x-axis? To display the lines a little closer together?

[grantmcdermott]

is it possible to have slightly larger margins/paddings at the left end right end of the x-axis? To display the lines a little closer together?

Not directly. Feel free to file a separate issue, though. (A workaround for the moment is too pass in a custom xlim value, e.g. c(0.5, 3.5).