Hi all,
I've been using the 'Ranger' random forest package alongside packages such as 'treeshap' to get Shapley values.
Yet, one thing I've noticed is that I am unable obtain the SHAP summary plot, typically known as the 'beeswarm' plot by using this package (or any random forest Shapley packages I could find online).
I wondered therefore if there was an equivalent package for random forest modelling to something like 'SHAPforXGboost' in which the beeswarm/summary plot was obtainable?
Any equivalent code equivalent to XGboost that would be suitable for Ranger would be greatly appreciated.
SHAP Visualization in R (first post) | Welcome to my blog
library(SHAPforxgboost)
library(xgboost)
library(data.table)
library(ggplot2)
})
X1 = as.matrix(iris[,-5])
mod1 = xgboost::xgboost(
data = X1, label = iris$Species, gamma = 0, eta = 1,
lambda = 0,nrounds = 1, verbose = F)
# shap.values(model, X_dataset) returns the SHAP
# data matrix and ranked features by mean|SHAP|
shap_values <- shap.values(xgb_model = mod1, X_train = X1)
shap_values$mean_shap_score
shap_values_iris <- shap_values$shap_score
# shap.prep() returns the long-format SHAP data from either model or
shap_long_iris <- shap.prep(xgb_model = mod1, X_train = X1)
# is the same as: using given shap_contrib
shap_long_iris <- shap.prep(shap_contrib = shap_values_iris, X_train = X1)
shap.plot.summary(shap_long_iris)