Convert a model to a skeleton — model

A "skeleton" is a data frame containing "nicely spaced" values for the explanatory variables in a model.

Usage

model_skeleton(mod, data = NULL, ncont = 3, nfirstcont = 50)

Arguments

mod: A fitted model or a tilde expression describing a model structure, e.g. outcome ~ vara+varb.
data: a data frame. Relevant only when mod is a tilde expression
ncont: minimum number of levels at which to represent continuous variables. (More levels may be added to "prettify" the choices. See pretty().)
nfirstcont: Like ncont, but for the first explanatory variable if it is categorical. This variable is mapped to the horizontal axis and so should have many levels to produce a smooth graph. (Default: 50)

Value

a data frame

Examples

Model <- FEV |> model_train(FEV ~ sex + age + height)
Model |> model_skeleton()
#>    sex age height
#> 1    F   0     40
#> 2    M   0     40
#> 3    F  10     40
#> 4    M  10     40
#> 5    F  20     40
#> 6    M  20     40
#> 7    F   0     50
#> 8    M   0     50
#> 9    F  10     50
#> 10   M  10     50
#> 11   F  20     50
#> 12   M  20     50
#> 13   F   0     60
#> 14   M   0     60
#> 15   F  10     60
#> 16   M  10     60
#> 17   F  20     60
#> 18   M  20     60
#> 19   F   0     70
#> 20   M   0     70
#> 21   F  10     70
#> 22   M  10     70
#> 23   F  20     70
#> 24   M  20     70
#> 25   F   0     80
#> 26   M   0     80
#> 27   F  10     80
#> 28   M  10     80
#> 29   F  20     80
#> 30   M  20     80