我在使用插入包rpart模型时遇到问题。当我在第1部分中运行代码时,出现以下错误:
library(rpart)
library(caret)
creditNames <- c("Checking", "Duration", "CreditHistory", "Purpose",
"CreditAmount", "Savings", "Employment",
"InstallmentRate", "GenderMarital", "OtherDebtors",
"YearsAtResidence", "RealEstate", "Age",
"OtherInstallment", "Housing", "ExistingCredit",
"Job", "NumLiable", "Phone", "Foreign", "Credit")
url="http://archive.ics.uci.edu/ml/machine-learning-databases/statlog/german/german.data"
credit_data <- read.table(url, sep=" ", header = FALSE,
col.names = creditNames,
stringsAsFactors = FALSE)
creditHistory <- c(A30="All Paid", A31="All Paid This Bank",
A32="Up To Date", A33="Late Payment",
A34="Critical Account")
credit_data$CreditHistory <- as.factor(creditHistory[credit_data$CreditHistory])
credit_data$Credit <- ifelse(credit_data$Credit == 1, "Good", "Bad")
credit_data$Credit <- factor(credit_data$Credit, levels = c("Good", "Bad"))
fitControl <- trainControl(method = 'cv', number = 6)
Grid <- expand.grid(
cp=.02)
vars <- names(credit_data)[c(5,13,3,7)]
samp.f <- as.formula(paste(names(credit_data)[21], paste(vars, collapse = " + "),sep="~"))
myvars <- c("Credit", "CreditAmount", "Age") #"CreditHistory", , "Employment"
samp.m <- train(samp.f,
data=credit_data,
method='rpart',
trControl = fitControl,
tuneLength=20,
metric = "Accuracy",
tuneGrid = Grid,
na.action = na.omit,
params=list(split='information'))
Fold1的模型拟合失败:cp = 0.02(函数(公式,数据, 权重,子集,na.action = na.rpart,:参数params not not 匹配
但是,当我在第2部分中运行代码时,我没有收到错误。
for (test_cp in seq(.001,.02,.001)){
test<-rpart(samp.f, dat=credit, cp=test_cp, maxdepth=5)
}
有什么建议吗?
我正在运行以下设置:
R version 3.4.4 (2018-03-15)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)
other attached packages:
[1] caret_6.0-79 ggplot2_2.2.1 lattice_0.20-35 rpart_4.1-13
答案 0 :(得分:2)
rpart
函数没有params选项,而是parms选项。所以在火车上你需要使用parms而不是params。
rpart(公式,数据,权重,子集,na.action = na.rpart,方法, model = FALSE,x = FALSE,y = TRUE, parms ,控制,成本,......)
samp.m <- train(samp.f,
data=credit_data,
method='rpart',
trControl = fitControl,
tuneLength=20,
metric = "Accuracy",
tuneGrid = Grid,
na.action = na.omit,
#--># parms=list(split='information'))