来一个data.table的写法


library(data.table)

c1 <- c('a', 2015, 300)
c2 <- c('a', 2016, 400)
c3 <- c('b', 2015, 700)
c4 <- c('b', 2016, 600)
dt <- data.frame(rbind(c1, c2, c3, c4), stringsAsFactors = FALSE)
names(dt) <- c('name', 'date', 'amount')

setDT(dt)

dt
#>    name date amount
#> 1:    a 2015    300
#> 2:    a 2016    400
#> 3:    b 2015    700
#> 4:    b 2016    600

# 一步到位
dt2 <- dt[,.SD[amount==max(amount),.(merged=paste0(date,"-",amount))],by=.(name)]

dt2
#>    name   merged
#> 1:    a 2016-400
#> 2:    b 2015-700

Created on 2020-01-02 by the reprex package (v0.3.0)

    dapengde

    结果貌似和我的不一样。

    你的结果是

    #>   name   merged
#> a    a 2015-300
#> b    b 2015-700

    a应该是2016-400

        借用上面的例子

        library(tidyverse)

c1 <- c('a', 2015, 300)
c2 <- c('a', 2016, 400)
c3 <- c('b', 2015, 700)
c4 <- c('b', 2016, 600)
dt <- data.frame(rbind(c1, c2, c3, c4), stringsAsFactors = FALSE)
names(dt) <- c('name', 'date', 'amount')

new.dt <- dt %>%
    group_by(name) %>%
    filter(amount == max(amount)) %>%
    mutate(merged = paste(date, amount, sep = "-")) %>%
    ungroup()

          dhfly 请教：继续这个例子，将其中一列拆分成2列怎么写？
          例：将id列分为q列和w列，以

          sep=“.”
ex=data.frame(id=c("100.a","002.b","003.c"),aumount=c(100,200,300))

          用 separate(data = ex, col =id, into = c("q", "w"), sep = ".")报错是什么原因？

              dhfly

              去掉sep = "."

              separate(data = ex, col =id, into = c("q", "w")) 即可

              如果你一定要手动指定，

              separate(data = ex, col =id, into = c("q", "w"),sep="\\.")

              报错是因为你传进去的sep会被当做一个正则表达式，.在正则表达式里match任何字符(except new line)。

              默认的sep会match任何non-alphanumeric 的字符，绝大多数情况下使用默认的即可。

              sep
              Separator between columns.

              If character, is interpreted as a regular expression. The default value is a regular expression that matches any sequence of non-alphanumeric values.

                  frankzhang21 非常感谢您的热情帮助，原来是这么回事，感觉自己不适合学语言。大白这里再谢！！

                    # 源数据
c1 <- c('a', 2015, 300)
c2 <- c('a', 2016, 400)
c3 <- c('b', 2015, 700)
c4 <- c('b', 2016, 600)
dt <- data.frame(rbind(c1, c2, c3, c4), stringsAsFactors = FALSE)
names(dt) <- c('name', 'date', 'amount')

library(dplyr)
library(tidyr)
dt %>% 
  group_by(name) %>% 
  filter(amount == max(amount)) %>% 
  unite("merged", date, amount, sep = "-") -> dt2
dt2