如何使用不同的变量以编程方式汇总tbl？

Question

我想用循环重写下面的代码来迭代by-variables＆＃39; a＆＃39;和＆＃39; b＆＃39;。我怎么能这样做？

set.seed(1234)
mydata <- as.tbl(data.frame(a=1:12, b=runif(12), x=rnorm(12)))
mydata

library(dplyr)
library(Hmisc)

mydata %>%
  mutate(feature=cut2(a, g=3)) %>% 
  group_by(feature) %>%  # group by quantiles of 'a'
  summarise(x=mean(x))

mydata %>%
  mutate(feature=cut2(b, g=3)) %>% 
  group_by(feature) %>%  # group by quantiles of 'b'
  summarise(x=mean(x))

Answer 1

我们可以使用interp中的lazyeval。

library(lazyeval)
lapply(c('a', 'b'), function(var)
                mydata %>% 
                  mutate_(feature=interp(~cut2(v1, g=3), v1=as.name(var))) %>% 
                          group_by(feature) %>%
                  summarise(x=mean(x)) )