其实hadley大人做了个实验性的dtplyr作为data.table的前端，干的事情是将dplyr的语法翻译到data.table，不过readme里写的也是心酸：都木有什么人用…
This is a large change that breaks all existing uses of dtplyr. But frankly, dtplyr was pretty useless before because it did such a bad job of generating data.table code. Fortunately few people used it, so a major overhaul possible.
我个人的理解是dplyr的性能应付日常操作足够了，真正在乎性能的用户都喜欢折腾，学个data.table也不是啥难事吧
Liechi data.table是不是有更简单的写法
library(data.table)
df <- as.data.table(mtcars)
df[2:5,cyl:=.(ifelse(cyl<6,2,cyl))]
library(magrittr)
library(dplyr)
mtcars[2:5, "cyl"] %<>% ifelse(.<6, 2, .)一年后，我用上了 data.table，写在这里给自己当做笔记，这是一个添加旬并且根据平闰年修改旬的日天数的代码：
time_seq <- qxun_sby[, c(1:3, 5)] %>%
.[, days := rep(c(10, 10, 11, 10, 10, 10, 10, 10, 11, 10, 10, 11, 10, 10, 10, 10, 10, 11,
10, 10, 10, 10, 10, 11, 10, 10, 11, 10, 10, 8, 10, 10, 11, 10, 10, 10),
times = iter / 36)] %>%
.[((y %% 100 == 0 & y %% 400 == 0) | (y %% 4 == 0 & y %% 100 != 0)) & t == 30, days := 9]