Add residualize_over_grid()

[strengejacke]

Fixes #129

[bwiernik]

Could we instead make this a model_residuals() function that takes a newdata argument, which could be a grid from get_datagrid()?

[strengejacke]

Sure, but do we need newdata? The grid is extracted / taken from the estimate_means() (or related) result.

[DominiqueMakowski]

It can still be useful for power-users to have a way to pass the data no?

[strengejacke]

@gemini-code-assist review

[strengejacke]

This is the current state - though I'm not sure how to fix the last error, and I don't understand why ggplot complains about discrete values?

[strengejacke]

library(modelbased)

model <- lm(Sepal.Length ~ Species * Sepal.Width, data = iris)

pr <- estimate_means(model, c("Sepal.Width", "Species"))
plot(pr, show_residuals = TRUE)

set.seed(1234)
x1 <- rnorm(200)
x2 <- rnorm(200)

# quadratic relationship
y <- 2 * x1 + x1^2 + 4 * x2 + rnorm(200)
d <- data.frame(x1, x2, y)

model <- lm(y ~ x1 + x2, data = d)

pr <- estimate_means(model, "x1")
plot(pr, show_residuals = TRUE)

model <- lm(y ~ x1 + x2, data = d)

pr <- estimate_means(model, c("x1", "x2=[sd]"))
plot(pr, show_residuals = TRUE)
#> Error in `scale_colour_continuous()`:
#> ! Discrete value supplied to a continuous scale.
#> ℹ Example values: -0.93, 0.07, and 1.08.

^{Created on 2025-08-16 with reprex v2.1.1}

[strengejacke]

It can still be useful for power-users to have a way to pass the data no?

I'm not quite sure at which point we should / can add newdata, since the data frame method already accepts a grid? Feel free to modify/add this function, where necessary.

[marcora]

Is there a way to get a table of partial residuals as easily as it is done here for plotting?

[strengejacke]

Yes, should be residualize_over_grid().

[marcora]

Thanks!!!

…

On Tue, Oct 14, 2025 at 1:28 PM Daniel ***@***.***> wrote: *strengejacke* left a comment (easystats/modelbased#386) <#386 (comment)> Yes, should be residualize_over_grid(). — Reply to this email directly, view it on GitHub <#386 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAABRR5E5SGIV6W2SHJVHFD3XUXFDAVCNFSM6AAAAABWZE34GOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIMBSHEZDCNRRGQ> . You are receiving this because you commented.Message ID: ***@***.***>