Question

尝试在已保存的rda模型上使用预测函数时，我有一个NullPointerException，请注意，当我直接使用该模型而不先保存它时，它会很好地工作，只有当我使用已保存的rda模型saveRDS()函数，然后使用readRDS()调用模型，我收到此错误。

完整错误：

Error in .jcall("RWekaInterfaces", "[D", "distributionForInstances", .jcast(classifier,  : 
  java.lang.NullPointerException

代码如下：

training_1<-read.table("/data/CHURN_FILES/training_1.txt", sep="|", header=TRUE)

Model<- train(
  x = subset(training_1, select = -c(Subs_Id,Churn_flag_OUT,Value_Segment,
                                     Global_Profile
  )), y = training_1$Churn_flag_OUT,
  metric = "ROC" ,
  method = "LMT",
  trControl = myControl)


saveRDS(Hichem_Model, "/data/CHURN_FILES/Model.rda")
Model_2 <- readRDS("/data/CHURN_FILES/Model.rda")

My_Data$predict <- predict(Model_2 ,(subset(My_Data, 
                                                           select = -c(Subs_Id,Churn_flag_OUT,Value_Segment,
                                                                       Global_Profile))) , type = "prob")[,"churn"]

在R中使用saveRDS（）时获取NullPointerException

0 个答案: