作者piggood (成为有品味的男人)
看板R_Language
标题Re: [问题] 用回圈整理同个ID的看病次数
时间Sun Jun 24 18:02:02 2018
※ 引述《chuubbyy (byetos)》之铭言：
: ※ 引述《piggood (成为有品味的男人)》之铭言：
: 我都比较建议用data.table的方式，简洁又快。
: library(data.table)
: setDT(medical)
: medical <- medical[,docvis.sum:=sum(docvis), by = "id"][,hhninc.mean:=mean(hhninc), by =
: "id"][, .SD[.N], by = "id"]
: 上面是data.table特殊chaining方式，若要考虑易读性可以分3行。
: medical[,docvis.sum:=sum(docvis), by = "id"]
: medical[,hhninc.mean:=mean(hhninc), by = "id"]
: medical <- medical[, .SD[.N], by = "id"]
: 或是pipe的方式。
: medical <- medical[,docvis.sum:=sum(docvis), by = "id"] %>%
: .[,hhninc.mean:=mean(hhninc), by = "id"] %>%
: .[, .SD[.N], by = "id"]
data.table 真的也不错用，功能强大速度也快，程式也精简
只是对我这个新手来说，还要再多多练习...
看懂 c 大的 code 之後，练习另一种写法，也可以达到原 po 的需求 ^^
medical <- medical[,':='(docvis.sum = sum(docvis), hhninc.mean = mean(hhninc)),
by ="id"][,.SD[.N], by = "id"]
如果本来的 hhninc 和 docvis 也要拿掉的话，後面再多加上一段
medical <- medical[,':='(docvis.sum = sum(docvis), hhninc.mean = mean(hhninc)),
by ="id"][,.SD[.N], by = "id"][,c("docvis","hhninc"):= NULL]
--
※ 发信站: 批踢踢实业坊(ptt.cc), 来自: 220.135.141.41
※ 文章网址: https://webptt.com/cn.aspx?n=bbs/R_Language/M.1529834526.A.359.html