1、如果调用了两个包，都有同样的函数，比如raster和dplyr都有select，就用dplyr::select
2、可以找出管道通道操作+contains、starts_with、ends_with
edu<-edu %>% select(contains("edu18"))
3、利用管道操作删样本
people<-sam[-which(sam$SAMPLEID%in%c("N130")),]
4、mutate操作（替换、新增变量超方便）
edu<-mutate(edu,EDU_UPDATE=ifelse(EDU_UPDATE==8,NA,EDU_UPDATE))
5、merge函数
edu<-merge(edu, place, by='SAMPLEID',all.x=TRUE)
6、!!、quo、unquo、!!!
详见vignette(“dplyr-programming”)
###两件事：（1）We need to quote the input ourselves；
（2）you use `!!` to say that you want to unquote an input so that it's evaluated, not quoted.
my_summarise <- function(df, group_var) {
df %>%group_by(!! group_var) %>% #!!你想要取消输入的引号，以便对其进行评估
summarise(a = mean(a))}
my_summarise(df, quo(g1))
###`!!!` takes a list of elements and splices them into to the current call.
args <- list(na.rm = TRUE, trim = 0.25)
quo(mean(x, !!! args))
args <- list(quo(x), na.rm = TRUE, trim = 0.25)
quo(mean(!!! args))
7、quo、unquo、!!!group分组+分组统计频数和百分比
fangyuan<-interviewer%>%group_by(编码)%>%summarise(n=n(),percentage=sum(反馈内容新)/sum(反馈内容总))
