在mlr中调优h2o超参数？

library(mlr)
library(h2o)

# 加载数据集
data(iris)

# 初始化h2o
h2o.init()

# 将数据转换为h2o格式
h2o_data <- as.h2o(iris)

# 定义超参数搜索空间
param_space <- list(
  learn_rate = makeNumericParam("learn_rate", lower = 0.01, upper = 0.1),
  max_depth = makeIntegerParam("max_depth", lower = 1, upper = 10)
)

# 定义评估指标
measure <- makeMeasure("classif.ce")

# 定义调优方法
ctrl <- makeTuneControlRandom(maxit = 100)

# 定义学习算法
learner <- makeLearner("classif.h2o.randomForest")

# 定义任务
task <- makeClassifTask(data = h2o_data, target = "Species")

# 运行调优过程
res <- tuneParams(learner = learner, task = task, resampling = cv3, measures = measure,
                  control = ctrl, par.set = param_space)

# 获取最佳超参数组合
best_params <- getTuneResult(res)$x

# 使用最佳超参数训练模型
model <- train(learner = learner, task = task, subset = res$opt.path[[1]]$x)