Let's support multiple models out of the box.
library(DALEX)
model1 <- glm(survived~., data = titanic_imputed, family = "binomial")
exp1 <- explain(model1, data = titanic_imputed)
pp1 <- predict_profile(exp1, titanic_imputed[1,])
library(ranger)
model2 <- ranger(survived~., data = titanic_imputed, probability = T)
exp2 <- explain(model2, titanic_imputed)
pp2 <- predict_profile(exp2, titanic_imputed[1,])
plot(pp1, pp2) # maybe should work automatically
plot(pp1, pp2, color = "_label_") # like this
Let's support multiple models out of the box.