# Load the necessary libraries
library(shapviz)
library(ggplot2)
library(xgboost)

# Read Data
dt <- read.csv(file.choose(), row.names = 1)
View(dt)
head(dt)
# Data Preparation
x <- dt[, 1:15]
y <- dt[, 16]

# Convert data to the DMatrix format required by XGBoost
dtrain <- xgb.DMatrix(data.matrix(x), label = y)

# Train the XGBoost model
fit <- xgb.train(
  params = list(learning_rate = 0.1, objective = "binary:logistic"),  
  data = dtrain,
  nrounds = 65L
)

# Generate SHAP explanation object
shp <- shapviz(fit, X_pred = data.matrix(x), X = x)

# Generate a standard SHAP feature importance plot
sv_importance(shp) +
  theme_bw() +
  theme(
    panel.grid.major = element_blank(),     
    panel.grid.minor = element_blank()      
  )

# Generate SHAP feature importance plots in the form of bar charts
sv_importance(shp, kind = "bar") +
  theme_bw() +
  theme(
    panel.grid.major = element_blank(),     
    panel.grid.minor = element_blank()      
  )

# Generate SHAP feature importance plots in the form of swarm plots and apply custom color schemes
sv_importance(shp, kind = "beeswarm") +
  theme_bw() +
  scale_color_gradientn(colors = c("#560B6A", "#FBB91E")) +  
  theme(
    legend.title = element_blank(),         
    panel.grid.major = element_blank(),     
    panel.grid.minor = element_blank(),   
    axis.line = element_line()              
  )