Question

我有一个像这样的数据框

ID <- c("ID21","ID22","ID23","ID24")
STR_PL_CAN_EVOLVE_PROCESS <- c("CCP_A,CCP_B","CCQ_A,CCQ_B,CCQ_C","IOT_A,CCP_B","CCQ_B,IOT_B")
Average <- c(7.5,6.5,7.1,6.6)
STR_VD_CAN_MEASURE_PROCESS <- c("Length,Breadth","Breadth,Width","Height,Length,Width","Width,Length")
Passfail <- c("Pass","Pass","Fail","Fail")

df <- data.frame(ID,STR_PL_CAN_EVOLVE_PROCESS,Average,STR_VD_CAN_MEASURE_PROCESS,Passfail,stringsAsFactors=FALSE)

我正在尝试使用tidyverse将以“ process”结尾的列中的值分成几列，并以此方式进行。

library(tidyverse)
df1 <- df %>%
  separate(STR_PL_CAN_EVOLVE_PROCESS, 
           paste0("ST_PL_CA_EV_PR","_Path",
                  seq(1:10)),
           sep = ",") %>%
  separate(STR_VD_CAN_MEASURE_PROCESS, 
           paste0("ST_VD_CA_ME_PR","_Path",
                  seq(1:10)),
           sep = ",")

这可行，但是我在这里手动做很多事情（输入列名，新列名）。这是我想要实现的一些事情

自动提供以“ 过程”结尾的列的名称，并将这些列分开。
提取列名中的前2个字符（用下划线分隔）作为新列名。例如： STR_PL_CAN_EVOLVE_PROCESS成为ST_PL_CA_EV_PR
删除仅不适用的列

我的所需输出是

    ID ST_PL_CA_EV_PR_Path1 ST_PL_CA_EV_PR_Path2 ST_PL_CA_EV_PR_Path3 Average ST_VD_CA_ME_PR_Path1 ST_VD_CA_ME_PR_Path2 ST_VD_CA_ME_PR_Path3 Passfail
  ID21                CCP_A                CCP_B                 <NA>     7.5               Length              Breadth                 <NA>     Pass
  ID22                CCQ_A                CCQ_B                CCQ_C     6.5              Breadth                Width                 <NA>     Pass
  ID23                IOT_A                CCP_B                 <NA>     7.1               Height               Length                Width     Fail
  ID24                CCQ_B                IOT_B                 <NA>     6.6                Width               Length                 <NA>     Fail

我的实际数据集中大约有35列以“ PROCESS”结尾。有人可以指出我正确的方向吗？

Answer 1

这里是cSplit

的一个选项

library(splitstackshape)
library(dplyr)
df %>% 
   cSplit(c("STR_PL_CAN_EVOLVE_PROCESS", "STR_VD_CAN_MEASURE_PROCESS"), 
                     ',', drop = TRUE)
#ID Average Passfail STR_PL_CAN_EVOLVE_PROCESS_1 STR_PL_CAN_EVOLVE_PROCESS_2 STR_PL_CAN_EVOLVE_PROCESS_3
#1: ID21     7.5     Pass                       CCP_A                       CCP_B                        <NA>
#2: ID22     6.5     Pass                       CCQ_A                       CCQ_B                       CCQ_C
#3: ID23     7.1     Fail                       IOT_A                       CCP_B                        <NA>
#4: ID24     6.6     Fail                       CCQ_B                       IOT_B                        <NA>
#   STR_VD_CAN_MEASURE_PROCESS_1 STR_VD_CAN_MEASURE_PROCESS_2 STR_VD_CAN_MEASURE_PROCESS_3
#1:                       Length                      Breadth                         <NA>
#2:                      Breadth                        Width                         <NA>
#3:                       Height                       Length                        Width
#4:                        Width                       Length                         <NA>

Answer 2

纯净的tidyverse版本。步骤很多，但我相信它会成就您的追求。

如何？

df %>% 

# Grabs all columns ending with process
gather(key,val, ends_with("PROCESS")) %>% 

# Separate all now previous column names by "_"
separate(key, paste0("Pat",1:5)) %>% 

# Mutate all columns starting with "Pat" to just two chars
mutate_at(vars(starts_with("Pat")), substr, 1, 2) %>% 

# Separat cell values on comma
separate(val, paste0("Path",1:3)) %>% 

# Gather all Path columns into a key and value pair
gather(Path, val,starts_with("Path")) %>% 

# Unite all columns starting with "Pat" into one
unite(key, starts_with("Pat")) %>% na.omit() %>% 

# Spread the data using "key" as columns, and "val" as vell values
spread(key,val)

    ID Average Passfail ST_PL_CA_EV_PR_Path1 ST_PL_CA_EV_PR_Path2 ST_PL_CA_EV_PR_Path3 ST_VD_CA_ME_PR_Path1
1 ID21     7.5     Pass                  CCP                    A                  CCP               Length
2 ID22     6.5     Pass                  CCQ                    A                  CCQ              Breadth
3 ID23     7.1     Fail                  IOT                    A                  CCP               Height
4 ID24     6.6     Fail                  CCQ                    B                  IOT                Width
  ST_VD_CA_ME_PR_Path2 ST_VD_CA_ME_PR_Path3
1              Breadth                 <NA>
2                Width                 <NA>
3               Length                Width
4               Length                 <NA>

将以“ Process”结尾的列拆分为R中具有各个值的许多列

2 个答案: