Error in estimte_slopes when trend is a factor

[DominiqueMakowski]

model <- lm(Sepal.Length ~ Species * Sepal.Width, data = iris)
modelbased::estimate_slopes(model, trend="Species", by="Sepal.Width")
#> Error in Summary.factor(structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, : 'range' not meaningful for factors

^{Created on 2025-01-09 with reprex v2.1.1}

[strengejacke]

It works with the marginaleffects backend, though, but I'm not sure which output is expected?

model <- lm(Sepal.Length ~ Species * Sepal.Width, data = iris)
modelbased::estimate_slopes(
  model,
  trend = "Species",
  by = "Sepal.Width",
  backend = "marginaleffects"
)
#> Estimated Marginal Effects
#> 
#> Parameter |          Comparison | Sepal.Width | Coefficient |   SE
#> ------------------------------------------------------------------
#> Species   | versicolor - setosa |        2.00 |        1.25 | 0.30
#> Species   |  virginica - setosa |        2.00 |        1.69 | 0.32
#> Species   | versicolor - setosa |        2.27 |        1.30 | 0.23
#> Species   |  virginica - setosa |        2.27 |        1.75 | 0.25
#> Species   | versicolor - setosa |        2.53 |        1.34 | 0.18
#> Species   |  virginica - setosa |        2.53 |        1.80 | 0.19
#> Species   | versicolor - setosa |        2.80 |        1.39 | 0.14
#> Species   |  virginica - setosa |        2.80 |        1.86 | 0.14
#> Species   | versicolor - setosa |        3.07 |        1.44 | 0.12
#> Species   |  virginica - setosa |        3.07 |        1.92 | 0.11
#> Species   | versicolor - setosa |        3.33 |        1.48 | 0.14
#> Species   |  virginica - setosa |        3.33 |        1.97 | 0.11
#> Species   | versicolor - setosa |        3.60 |        1.53 | 0.19
#> Species   |  virginica - setosa |        3.60 |        2.03 | 0.15
#> Species   | versicolor - setosa |        3.87 |        1.58 | 0.25
#> Species   |  virginica - setosa |        3.87 |        2.08 | 0.21
#> Species   | versicolor - setosa |        4.13 |        1.62 | 0.31
#> Species   |  virginica - setosa |        4.13 |        2.14 | 0.27
#> Species   | versicolor - setosa |        4.40 |        1.67 | 0.37
#> Species   |  virginica - setosa |        4.40 |        2.20 | 0.33
#> 
#> Parameter |       95% CI |     t |      p
#> -----------------------------------------
#> Species   | [0.67, 1.83] |  4.23 | < .001
#> Species   | [1.07, 2.31] |  5.35 | < .001
#> Species   | [0.84, 1.76] |  5.53 | < .001
#> Species   | [1.25, 2.24] |  6.92 | < .001
#> Species   | [0.99, 1.69] |  7.51 | < .001
#> Species   | [1.43, 2.18] |  9.36 | < .001
#> Species   | [1.12, 1.66] | 10.19 | < .001
#> Species   | [1.58, 2.13] | 13.24 | < .001
#> Species   | [1.20, 1.67] | 11.79 | < .001
#> Species   | [1.70, 2.13] | 17.74 | < .001
#> Species   | [1.20, 1.76] | 10.31 | < .001
#> Species   | [1.75, 2.19] | 17.37 | < .001
#> Species   | [1.16, 1.90] |  8.04 | < .001
#> Species   | [1.73, 2.33] | 13.25 | < .001
#> Species   | [1.09, 2.06] |  6.37 | < .001
#> Species   | [1.68, 2.49] | 10.02 | < .001
#> Species   | [1.02, 2.23] |  5.24 | < .001
#> Species   | [1.61, 2.67] |  7.95 | < .001
#> Species   | [0.94, 2.40] |  4.46 | < .001
#> Species   | [1.54, 2.85] |  6.60 | < .001
#> Marginal effects estimated for Species

^{Created on 2025-01-09 with reprex v2.1.1}

[DominiqueMakowski]

That seems correct, the expected output would be all the species-related parameters (i.e., virginica and versicolor (minus the ref) unfolded over Sepal.Width