Question

我有大数据集，前两列看起来与此相似：

team year

Arizona 2006

Arizona 2006

Arizona 2011

Oregon 2011

Oklahoma 2008

Colorado 2005

Colorado 2005

Colorado 2011

我想根据团队和年份创建团队所在会议的另一个变量。我在考虑类似if(data$team="Arizona|Oregon|Colorado" & year=2011){data$conf='Pac-12}之类的东西，但这并不起作用，因为通常会有多行拥有团队和我正在寻找的年份。这有意义吗？

谢谢！

Answer 1

也许你需要ifelse？

teams <- c("Arizona", "Oregon", "Colorado")

data$conf <- ifelse(data$team %in%  teams & data$year == 2011,
                     "Pac-12", "something else")

修改的

您可以通过子集化来更改内容，因为@Simplefish已经向您显示了您是否不想要if。虽然你的问题确实要求if。

另一种方法是一次完成所有更改，因此您不要过度复制答案。您可以将if嵌套为：

ifelse(data$team %in%  teams & data$year == 2011,
       "Pac-12", ifelse(data$team %in% "Oklahoma" & data$year == 2008,
                 "second answer", "third answer"))

但是在很多条件下这很麻烦，所以也许你想要：

reference <- matrix(c(rep("Pac-12",3),rep("third answer",4),
                     "Second Answer",rep("fourth answer",8)), 
                    4, 4, 
                    dimnames=list(c("Arizona","Oregon", "Colorado", "Oklahoma"), 
                                  c("2011","2008","2006","2005") )
                    )

#> reference
#         2011           2008            2006            2005           
#Arizona  "Pac-12"       "third answer"  "fourth answer" "fourth answer"
#Oregon   "Pac-12"       "third answer"  "fourth answer" "fourth answer"
#Colorado "Pac-12"       "third answer"  "fourth answer" "fourth answer"
#Oklahoma "third answer" "Second Answer" "fourth answer" "fourth answer"


data$conf <- with( data, reference [ cbind(team,year) ] )
#    > data
#      team year          conf
#1  Arizona 2006 fourth answer
#2  Arizona 2006 fourth answer
#3  Arizona 2011        Pac-12
#4   Oregon 2011        Pac-12
#5 Oklahoma 2008 Second Answer
#6 Colorado 2005 fourth answer
#7 Colorado 2005 fourth answer
#8 Colorado 2011        Pac-12

最后一种方法是merge使用data.frame版本的引用.....我相信其他人可能会证明这一点。

Answer 2

或者你可以通过直接子集来实现：

data$conf <- rep(NA,nrow(data))
data$conf[(data$team == 'Arizona' | data$team == 'Oregon' | data$team == 'Colorado') & data$year == 2011]='PAC-12'

基于R数据帧中的两个其他变量替换字符值

2 个答案: