使用“交互”package 在 R 中创建多面图 2 向交互

Question

I am using the cat_plot function from the 'interactions' package in R (which is a wrapper for ggplot) to plot a 2-way interaction with 2 categorical variables. I can do this easily using the code below (reprex from the "diamonds" dataset)

require(interactions)

data("diamonds")

m <- glm(price ~ cut*color, data = diamonds)

cat_plot(m, pred = cut, modx = color, geom = "bar", colors = "Set1")

This produces the following graph

However, what I would like is to have a faceted graph with each of the cuts presented separately, to make it visually easier to interpret. This can be done for 3-way interactions using the facet.modx = TRUE command, but when I try this with only a 2-way interaction with cat_plot(m, pred = cut, modx = color, geom = "bar", colors = "Set1", facet.modx = TRUE) I get the following error

Error in prep_data(model = model, pred = pred, modx = modx, pred.values = pred.values,  : 
  formal argument "facet.modx" matched by multiple actual arguments

Is there a way to easily facet the graph for 2 way interactions? My real-life dataset is actually a glmer model so I would prefer to stay within the "interactions" package if possible.

EDIT: based on the suggestion from @stefan I tried the following syntax cat_plot(m, pred = cut, modx = color, geom = "bar", colors = "Set1") + facet_wrap(~cut) which produced the graph below. This is almost exactly what I want, except it has seemed to keep the other 'cuts' on the x-axis and just removed the bars. Ideally, colours would be on the x-axis instead.

EDIT 2:

I have recreated the problem using data which is more similar to what I am actually working with, with a binary outcome, random effects from glmer etc.

require(lme4)
require(interactions)

set.seed(123)
id <- rep(1:150, each = 4)
condition <- rep(c("a", "b", "c"), each = 4, times = 50)
cat_mod <- rep(c("cat_1", "cat_2", "cat_3", "cat_4"), each = 1, length.out = 600)
control_mod <- rep(c("control_1", "control_2"), each = 4, length.out = 600)
binary_choice <- rbinom(600, 1, 0.5)


simdat <- data.frame(id, condition, cat_mod, binary_choice, control_mod)

m <- glmer(binary_choice ~ condition*cat_mod + control_mod + (1 | id), family=binomial, data = simdat)

cat_plot(m, pred = condition, modx = cat_mod, geom = "bar", colors = "Set1")

I would like to preserve the response scale on the y-axis, and the model accounting for the random intercept, which is why I was trying to avoid using ggplot directly, as the interactions package is already built to accommodate glmms, which is super convenient.

SOLVED

Following the suggestion from @RStam I modified the code slightly so that all y-axes had the same scale, and removed the duplicate facet labels at the bottom.

cat_plot(m, pred = condition, modx = cat_mod, geom = "bar", colors = "Set1") + 
  scale_x_discrete(labels = c(a = " ", b = " ", c = " ")) + 
  facet_wrap(condition~., scales= "free_x")

This was the final result

Answer 1

Original Answer

cat_plot(m, pred = cut, modx = color, geom = "bar", colors = "Set1") + 
facet_wrap(~cut, scales = "free_x")

Edit 1

After that it still wasn't resolving your issue I've updated my answer. This should resolve the issue you are having.

library(tidyverse)
ggplot(diamonds, aes(x=color,y=price, fill = color)) +
geom_col() + facet_wrap(~cut, scales = "free")

Edit 2

Using your new data and the interactions package I found a rather unpleasant 'hack' using scale_x_discrete() but it should give the desired outcome.

library(interactions)
library(lme4)

set.seed(123)
id <- rep(1:150, each = 4)
condition <- rep(c("a", "b", "c"), each = 4, times = 50)
cat_mod <- rep(c("cat_1", "cat_2", "cat_3", "cat_4"), each =  
1, length.out = 600)
control_mod <- rep(c("control_1", "control_2"), each = 4, 
length.out = 600)
binary_choice <- rbinom(600, 1, 0.5)


simdat <- data.frame(id, condition, cat_mod, binary_choice, 
control_mod)

m <- glmer(binary_choice ~ condition*cat_mod + control_mod + 
(1 | id), family=binomial, data = simdat)

cat_plot(m, pred = condition, modx = cat_mod, geom = "bar", 
colors = "Set1") + scale_x_discrete() + 
facet_wrap(condition~., scales= "free")