Plot MID with ggplot2 Package

For "mid" objects, ggmid() visualizes a MID component function using the ggplot2 package.

Usage

ggmid(object, ...)

# S3 method for class 'mid'
ggmid(
  object,
  term,
  type = c("effect", "data", "compound"),
  theme = NULL,
  intercept = FALSE,
  main.effects = FALSE,
  data = NULL,
  jitter = 0.3,
  cells.count = c(100L, 100L),
  limits = c(NA, NA),
  ...
)

# S3 method for class 'mid'
autoplot(object, ...)

Arguments

object: a "mid" object to be visualized.
...: optional parameters to be passed to the main layer.
term: a character string specifying the component function to be plotted.
type: character string. The method for plotting the interaction effects.
theme: a character string specifying the color theme or any item that can be used to define "color.theme" object.
intercept: logical. If TRUE, the intercept is added to the MID values.
main.effects: logical. If TRUE, the main effects are included in the interaction plot.
data: a data.frame to be plotted with the corresponding MID values. If not passed, data is extracted from parent.env() based on the function call of the "mid" object.
jitter: a numeric value specifying the amount of jitter for points.
cells.count: an integer or integer-valued vector of length two, specifying the number of cells for the raster type interaction plot.
limits: NULL or a numeric vector of length two specifying the limits of the plotting scale. NAs are replaced by the minimum and/or maximum MID values.

Value

ggmid.mid() returns a "ggplot" object.

Details

The S3 method of ggmid() for "mid" objects creates a "ggplot" object that visualizes a MID component function. The main layer is drawn using geom_line() or geom_path() for a main effect of a quantitative variable, geom_col() for a main effect of a qualitative variable, and geom_raster() or geom_rect() for an interaction effect. For other methods of ggmid(), see help(ggmid.mid.importance), help(ggmid.mid.breakdown) or help(ggmid.mid.conditional).

Examples

data(diamonds, package = "ggplot2")
set.seed(42)
idx <- sample(nrow(diamonds), 1e4)
mid <- interpret(price ~ (carat + cut + color + clarity)^2, diamonds[idx, ])
#> 'model' not passed: response variable in 'data' is used
ggmid(mid, "carat")

ggmid(mid, "clarity")

ggmid(mid, "carat:clarity", main.effects = TRUE)

ggmid(mid, "clarity:color", type = "data", theme = "Mako", data = diamonds[idx, ])

ggmid(mid, "carat:color", type = "compound", data = diamonds[idx, ])