通过两列连接进行子集化

我想同时在两列上创建数据调节的子集。通过两列连接进行子集化

类似于此： subsetting data using multiple variables in R

例如：

说我有这个叫Gamedat数据集：

 Games People Hoursplayed 
    goldeneye Michael   5 
    goldeneye Thatcher   8 
    goldeneye Dexter   12 
    goldeneye Dexter   15 
     pacman Dexter   2 
     tetris  Clint   5 
     tetris Dexter   8 
    goldeneye Thatcher   12 
     pacman Thatcher   15 
    goldeneye  Clint   2 
     pacman Michael   5 
     pacman Michael   8 
     pacman  Clint   12 
     tetris  John   15 
     tetris  Clint   2 
ageofempires  Clint   5 
     pacman Dexter   8 
ageofempires Thatcher   12 
ageofempires  John   15 
    goldeneye Dexter   2

说我想看看像鹊游戏。我想看看任何玩家玩过其他游戏的时间与他们玩过goldeneye的时间相同（这在我的真实数据集中更有用）。

所以我这样做：

Gameofinterest <- Gamedat[ grep("goldeneye", Gamedat[ ,1]), ]`

那么我这样做：

subset(Gamedat, Gamedat[ ,2] %in% Gameofinterest[ ,2] & 
    Gamedat[ ,3] %in% Gameofinterest[ ,3])

但是这给了我：

 Games People Hoursplayed 
    goldeneye Michael   5 
    goldeneye Thatcher   8 
    goldeneye Dexter   12 
    goldeneye Dexter   15 
     pacman Dexter   2 
     tetris Clint   5 
     tetris Dexter   8 
    goldeneye Thatcher   12 
     pacman Thatcher   15 
    goldeneye Clint   2 
     pacman Michael   5 
     pacman Michael   8 
     pacman Clint   12 
     tetris Clint   2 
ageofempires Clint   5 
     pacman Dexter   8 
ageofempires Thatcher   12 
    goldeneye Dexter   2

当我真正想要的是这样的：

  Games People Hoursplayed 
    goldeneye Michael   5 
    goldeneye Thatcher   8 
    goldeneye Dexter   12 
    goldeneye Dexter   15 
     pacman Dexter   2 
    goldeneye Thatcher   12 
    goldeneye Clint   2 
     pacman Michael   5 
     tetris Clint   2 
    ageofempires Thatcher   12 
    goldeneye Dexter   2

总之，我要找到匹配的“人& Hoursplayed”那个例子，

，而不是“人” &“Hoursplayed” ...有意义吗？

我知道我能做到这一点：

Gamedat$PHpaste <- paste(Gamedat$People, Gamedat$Hoursplayed, sep="") 

Gamedat[Gamedat[ ,4] %in% Gameofinterest[ ,4], ]

，并得到：

 Games People Hoursplayed PHpaste 
    goldeneye Michael   5 Michael5 
    goldeneye Thatcher   8 Thatcher8 
    goldeneye Dexter   12 Dexter12 
    goldeneye Dexter   15 Dexter15 
     pacman Dexter   2 Dexter2 
    goldeneye Thatcher   12 Thatcher12 
    goldeneye Clint   2  Clint2 
     pacman Michael   5 Michael5 
     tetris Clint   2  Clint2 
ageofempires Thatcher   12 Thatcher12 
    goldeneye Dexter   2 Dexter2

希望的东西更优雅？

来源

2017-02-28 StatGenGeek

是您期望的结果是否正确？德克斯特已经打了2个小时的pacman，但是打了29个小时的goldeneye ......是不是因为这29个小时中有2个是独特记录的一部分？ – shayaa

最后一行显示德克斯特已经玩了2个小时，所以这是一个正确的比赛。 – StatGenGeek

我认为这可以使用dplyr来实现。首先，使用过滤器检索游戏是否是goldeneye的行。然后使用inner_join使用People和HoursPlayed加入原始数据。可选：选择所需的列并按人员排列。

library(dplyr) 
Gamedat %>% 
    filter(Games == "goldeneye") %>% 
    inner_join(Gamedat, by = c("People", "Hoursplayed")) %>% 
    select(Games = Games.y, People, Hoursplayed) %>% 
    arrange(People)

结果：

  Games People Hoursplayed 
1  goldeneye Clint   2 
2  tetris Clint   2 
3  goldeneye Dexter   12 
4  goldeneye Dexter   15 
5  pacman Dexter   2 
6  goldeneye Dexter   2 
7  goldeneye Michael   5 
8  pacman Michael   5 
9  goldeneye Thatcher   8 
10 goldeneye Thatcher   12 
11 ageofempires Thatcher   12

来源

2017-02-28 03:30:08 neilfws

美丽的谢谢你。 – StatGenGeek

通过两列连接进行子集化

回答

相关问题