mice填补

library(readxl)
nv188 <- read_excel("C:/Users/Lenovo/Desktop/nv188.xlsx")
View(nv188)
library(mice)
mice_mod<-mice(data=nv188,m=5,method=c("pmm"),maxit=5,seed = 1234)
mice_output<-complete(mice_mod)
View(mice_output)
data6<-complete(mice_mod,action = 1)
View(data6)
data7<-complete(mice_mod,action = 2)
View(data7)
data8<-complete(mice_mod,action = 3)
View(data8)
data9<-complete(mice_mod,action = 4)
View(data9)
data10<-complete(mice_mod,action = 5)
View(data10)
write.csv(data10,"data10.csv")
write.csv(data10,"data9.csv")
write.csv(data10,"data8.csv")
write.csv(data10,"data7.csv")
write.csv(data10,"data6.csv")

缺失数据集——MCMC估计插补成几个数据集——每个数据集进行插补建模（glm、lm模型）——将这些模型整合到一起（pool）——评价插补模型优劣（模型系数的t统计量）——输出完整数据集（compute）

MCAR