R - 如何重新整理两列中的数据帧整理值?
问题描述:
我有一个数据框,我需要重塑,以方便它在一个viz应用程序中使用。这里的数据帧的压缩版本:R - 如何重新整理两列中的数据帧整理值?
Carrier <- c("Mesa", "United", "JetBlue", "ExpressJet", "SkyWest")
Flight_Num <- c(7124, 7177, 334, 1223, 6380)
Origin <- c("ORD", "EWR", "SFO", "BOS", "BDL")
Dest <- c("PIT", "BOI", "DSM", "CWA", "CMH")
Sched_Depr <- c(1955, 1900, 1845, 1253, 1755)
df <- data.frame(Carrier, Flight_Num, Origin, Dest, Sched_Depr)
Carrier Flight_Num Origin Dest Sched_Depr
1 Mesa 7124 ORD PIT 1955
2 United 7177 EWR BOI 1900
3 JetBlue 334 SFO DSM 1845
4 ExpressJet 1223 BOS CWA 1253
5 SkyWest 6380 BDL CMH 1755
Origin
和Dept
被解释为地理数据(即坐标)由应用程序即。我需要将它们整理在一个名为Coords
的列中。同时我需要创建一个新的订单标记变量Order_Points
。因此,新的,重塑数据帧应该是这样的:
Carrier Flight_Num Coords Sched_Depr Order_Points
1 Mesa 7124 ORD 1955 1
2 Mesa 7124 PIT 1955 2
3 United 7177 EWR 1900 1
4 United 7177 BOI 1900 2
5 JetBlue 334 SFO 1845 1
6 JetBlue 334 DSM 1845 2
7 ExpressJet 1223 BOS 1253 1
8 ExpressJet 1223 CWA 1253 2
9 SkyWest 6380 BDL 1755 1
10 SkyWest 6380 CMH 1755 2
什么是整理两列这样一种有效的方式,同时保持(和复制)其他变量?
答
这里有一个选项使用tidyverse函数。我们使用gather
将数据帧从“宽”格式转换为“长格式”。这还添加了一个列(在此称为Type
),标记Coords
是Origin
还是Dest
。
library(tidyverse)
df.long = df %>%
gather(Type, Coords, Origin, Dest) %>%
arrange(Carrier, desc(Type))
Carrier Flight_Num Sched_Depr Type Coords 1 ExpressJet 1223 1253 Origin BOS 2 ExpressJet 1223 1253 Dest CWA 3 JetBlue 334 1845 Origin SFO 4 JetBlue 334 1845 Dest DSM 5 Mesa 7124 1955 Origin ORD 6 Mesa 7124 1955 Dest PIT 7 SkyWest 6380 1755 Origin BDL 8 SkyWest 6380 1755 Dest CMH 9 United 7177 1900 Origin EWR 10 United 7177 1900 Dest BOI
答
您还可以使用基础R:
dat <- data.frame(Carrier, Flight_Num, Origin, Dest, Sched_Depr)
df=reshape(dat,idvar = "Carrier",varying = list(3:4),direction = "long")
`row.names<-`(df[order(df[,1]),],NULL)
Carrier Flight_Num Sched_Depr time Origin
1 ExpressJet 1223 1253 1 BOS
2 ExpressJet 1223 1253 2 CWA
3 JetBlue 334 1845 1 SFO
4 JetBlue 334 1845 2 DSM
5 Mesa 7124 1955 1 ORD
6 Mesa 7124 1955 2 PIT
7 SkyWest 6380 1755 1 BDL
8 SkyWest 6380 1755 2 CMH
9 United 7177 1900 1 EWR
10 United 7177 1900 2 BOI
可以时间变量名称更改为你喜欢在你上面的例子
轻微的一个改进:'df%>% 聚集(Order_Points,Coords,Origin,Dest)%>% mutate(Order_Points => ifelse(Order_Points ==“Origin”,1,2))%>% select(Carrier,Flight_Num,Coords,Sched_Depr,Order_Points)%>% arrange(desc(Sched_Depr))' – jsb