cv

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
nround <- 1500
param <- list(max_depth=9, eta=0.01 , silent=1, nthread=4, objective='binary:logistic')

model_data = data_higgs_0_cleaned_scaled
label = as.integer(model_data$Label) -1
drop = c(ncol(model_data),ncol(model_data)-1,ncol(model_data)-2)
dtrain <- xgb.DMatrix(data.matrix(model_data[,-drop]),label = label)

cv = xgb.cv(param, dtrain, nround, nfold=5, metrics={'auc'})
plot(cv$evaluation_log$iter,1-cv$evaluation_log$test_auc_mean)

cv_0 = cv
ntrees_0 = which((1-cv$evaluation_log$test_auc_mean) == min(1-cv$evaluation_log$test_auc_mean))
ntrees_0
# 624

model_data = data_higgs_1_cleaned_scaled

label = as.integer(model_data$Label) -1
drop = c(ncol(model_data),ncol(model_data)-1,ncol(model_data)-2)
dtrain <- xgb.DMatrix(data.matrix(model_data[,-drop]),label = label)
cv = xgb.cv(param, dtrain, nround, nfold=5, metrics={'auc'})
plot(cv$evaluation_log$iter,1-cv$evaluation_log$test_auc_mean)


cv_1     = cv
ntrees_1 = which((1-cv$evaluation_log$test_auc_mean) == min(1-cv$evaluation_log$test_auc_mean))
ntrees_1
#  652


model_data = data_higgs_2_cleaned_scaled

label = as.integer(model_data$Label) -1
drop = c(ncol(model_data),ncol(model_data)-1,ncol(model_data)-2)
dtrain <- xgb.DMatrix(data.matrix(model_data[,-drop]),label = label)
cv = xgb.cv(param, dtrain, nround, nfold=5, metrics={'auc'})
plot(cv$evaluation_log$iter,1-cv$evaluation_log$test_auc_mean)

ntrees_2 = which((1-cv$evaluation_log$test_auc_mean) == min(1-cv$evaluation_log$test_auc_mean))

cv_2     = cv

ntrees_2
# 1104

#############################################################