Caret train 方法在尝试使用 H2o 包时抱怨：“出了点问题；所有准确度指标值都丢失了”答案

【问题标题】：Caret train method complains when trying to use H2o package: "Something is wrong; all the Accuracy metric values are missing"Caret train 方法在尝试使用 H2o 包时抱怨：“出了点问题；所有准确度指标值都丢失了”
【发布时间】：2020-11-30 05:09:18
【问题描述】：

当我想在此示例中使用插入符号中的 H2o 方法时收到错误消息：

library(caret)
library(h2o)

data(HELPrct)
ds = HELPrct
fitControl= trainControl(method="repeatedcv", number = 5)
ds$sub = as.factor(ds$substance)
h2oFit1 <- train(homeless ~ female + i1 + sub + sexrisk + mcs + pcs, 
               trControl=fitControl, 
               method = "gbm_h2o", 
               data=ds[complete.cases(ds),])

然后R告诉我：

Something is wrong; all the Accuracy metric values are missing:
   Accuracy       Kappa    
 Min.   : NA   Min.   : NA
 1st Qu.: NA   1st Qu.: NA
 Median : NA   Median : NA
 Mean   :NaN   Mean   :NaN 
 3rd Qu.: NA   3rd Qu.: NA
 Max.   : NA   Max.   : NA  
 NA's   :9     NA's   :9 
Error: Stopping
In addition: Warning message:
In nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo,  :
  There were missing values in resampled performance measures.

有谁知道我如何使插入符号与 h2o 一起工作？其他方法都没有问题。

【问题讨论】：

标签： r r-caret h2o

【解决方案1】：

如果我运行您的代码并检查警告：

warnings()
Warning messages:
1: model fit failed for Fold1.Rep1: max_depth=1, ntrees= 50, learn_rate=0.1, min_rows=10, col_sample_rate=1 Error in h2o.getConnection() : 
  No active connection to an H2O cluster. Did you run `h2o.init()` ?

所以你需要h2o.init()（查看webpage了解更多详情）：

library(caret)
library(h2o)

h2o.init()
ds = mosaicData::HELPrct

fitControl= trainControl(method="repeatedcv", number = 5)
ds$sub = as.factor(ds$substance)
h2oFit1 <- train(homeless ~ female + i1 + sub + sexrisk + mcs + pcs, 
               trControl=fitControl, 
               method = "gbm_h2o", 
               data=ds[complete.cases(ds),])

h2oFit1
Gradient Boosting Machines 

117 samples
  6 predictor
  2 classes: 'homeless', 'housed' 

No pre-processing
Resampling: Cross-Validated (5 fold, repeated 1 times) 
Summary of sample sizes: 93, 94, 93, 94, 94 
Resampling results across tuning parameters:

  max_depth  ntrees  Accuracy   Kappa    
  1           50     0.5826087  0.0669072
  1          100     0.6253623  0.1895957
  1          150     0.6420290  0.2188447
  2           50     0.6159420  0.1708235
  2          100     0.6072464  0.1513658
  2          150     0.6329710  0.2035319
  3           50     0.6253623  0.1878658
  3          100     0.6159420  0.1701928
  3          150     0.6420290  0.1761487

【讨论】：