我正在尝试对R中的所有列进行ChISQUARE测试。

我的数据:

dependent <- c(1,0,1)
Salutation <- c("MR","MR","MR")
Gender <- c("MALE","MALE","MALE")
Profession <- c("BUSINESS","OTHER","BUSINESS")
Region <- c("South","West","Midwest")
CUST_RES_CITY <- c("MUMBAI","BANGALORE","BANGALORE")
Vertical_New <- c("AGENCY 1","AGENCY 1","AGENCY 2")
VehicleMake  <-c ("SUZUKI","HONDA","VOLKS") 
model<- c("SWIFT","MOBILIO","POLO")
Segment <- c("5 to 10 Lacs","10 to 30 Lacs","5 to 10 Lacs")


factor.cs <- data.frame(dependent, Salutation ,Gender ,Profession ,Region, CUST_RES_CITY,Vertical_New,VehicleMake ,model,Segment)

以前我有这一块可以做卡方检验,并会从中返回统计数据。

corr.factor <- factor.cs %>%
summarise_each(funs(chisq.test(., 
                               factor.cs[,dependent])$statistic), -one_of(dependent))

现在,由于summary_each被改为summary_at,_all等,所以我试图修改代码

corr.factor <- factor.cs %>%  
 summarise_all(funs(chisq.test(., factor.cs$dependent)$statistic))
 %>% select(-c("dependent"))

我得到这个错误:

Error in summarise_impl(.data, dots) : 
  Evaluation error: 'x' and 'y' must have at least 2 levels