R 用数据帧中的列名替换1s

我的数据如下所示：

df<-data.frame(
 hhid = c(5668,5595,4724,4756,4856,4730,4757,6320,4758,6319,6311,5477,6322),

 pid=c(93660,93660,100960,100960,100960,100960,100960,100962,100960,100962,100962,93814,100962),

 pname=c("AB","HG","DC","DC","DC","DC","DC","BA","DC","BA","BA","BH","BA"),

 vid=c(462962,608757,772284,772284,772284,293475,293475,656323,293475,656323,81533,465612,656323),

 vname=c("ABCD","KJUI","HIND","HIND","HIND","KAJA","KAJA","ADAR","KAJA","ADAR","SANG","NARI","ADAR"),

 me=c(1,1,0,0,0,0,0,1,0,0,0,1,0),

 ls=c(0,0,1,1,0,1,0,0,1,0,0,0,1),

 lg=c(0,0,0,0,0,0,1,0,0,1,0,0,0),

 lo=c(0,0,0,0,1,0,0,0,0,0,1,0,0),

 amt=c(20000,20000,14000,14000,14000,14000,14000,27000,14000,27000,27000,20000,23000))

---

因此，基本上我想要的是用一个新的列LOS替换me、ls、lg、lo列中的1，该列LOS包含列名me、ls、lg、lo作为新行数据。这就是用列名替换1s，并将其保存到一个新的列LOS。

您可以将数据转换为更长的格式，并

过滤
值为1的行。

library(dplyr)

tidyr::pivot_longer(df, cols = c(me, ls, lg, lo), names_to = "LOS") %>%
  filter(value == 1) %>%
  select(-value)

#   hhid    pid pname    vid vname   amt LOS 
#   <dbl>  <dbl> <fct>  <dbl> <fct> <dbl> <chr>
# 1  5668  93660 AB    462962 ABCD  20000 me   
# 2  5595  93660 HG    608757 KJUI  20000 me   
# 3  4724 100960 DC    772284 HIND  14000 ls   
# 4  4756 100960 DC    772284 HIND  14000 ls   
# 5  4856 100960 DC    772284 HIND  14000 lo   
# 6  4730 100960 DC    293475 KAJA  14000 ls   
# 7  4757 100960 DC    293475 KAJA  14000 lg   
# 8  6320 100962 BA    656323 ADAR  27000 me   
# 9  4758 100960 DC    293475 KAJA  14000 ls   
#10  6319 100962 BA    656323 ADAR  27000 lg   
#11  6311 100962 BA     81533 SANG  27000 lo   
#12  5477  93814 BH    465612 NARI  20000 me   
#13  6322 100962 BA    656323 ADAR  23000 ls   

库（dplyr）
tidyr:：pivot_更长（df，cols=c（me，ls，lg，lo），名称_to=“LOS”）%>%
过滤器（值==1）%>%
选择（-value）
#hhid pid pname vid vname amt LOS
#           
#15668 93660 AB 462962 ABCD 20000 me
#2559593660汞柱608757 KJUI 20000 me
#34724100960 DC 772284后14000 ls
#4 4756 100960 DC 772284后14000 ls
#54856100960直流772284后14000低
#6 4730 100960 DC 293475 KAJA 14000 ls
#7 4757 100960 DC 293475 KAJA 14000 lg
#8 6320 100962 BA 656323 ADAR 27000 me
#9 4758 100960 DC 293475 KAJA 14000 ls
#10631100962 BA 656323阿达尔27000 lg
#116311100962BA81533SANG27000LO
#12547793814 BH 465612 NARI 20000 me
#13 6322 100962 BA 656323 ADAR 23000 ls
您可以将数据转换为更长的格式，并
过滤
值为1的行

library(dplyr)

tidyr::pivot_longer(df, cols = c(me, ls, lg, lo), names_to = "LOS") %>%
  filter(value == 1) %>%
  select(-value)

#   hhid    pid pname    vid vname   amt LOS 
#   <dbl>  <dbl> <fct>  <dbl> <fct> <dbl> <chr>
# 1  5668  93660 AB    462962 ABCD  20000 me   
# 2  5595  93660 HG    608757 KJUI  20000 me   
# 3  4724 100960 DC    772284 HIND  14000 ls   
# 4  4756 100960 DC    772284 HIND  14000 ls   
# 5  4856 100960 DC    772284 HIND  14000 lo   
# 6  4730 100960 DC    293475 KAJA  14000 ls   
# 7  4757 100960 DC    293475 KAJA  14000 lg   
# 8  6320 100962 BA    656323 ADAR  27000 me   
# 9  4758 100960 DC    293475 KAJA  14000 ls   
#10  6319 100962 BA    656323 ADAR  27000 lg   
#11  6311 100962 BA     81533 SANG  27000 lo   
#12  5477  93814 BH    465612 NARI  20000 me   
#13  6322 100962 BA    656323 ADAR  23000 ls   

库（dplyr）
tidyr:：pivot_更长（df，cols=c（me，ls，lg，lo），名称_to=“LOS”）%>%
过滤器（值==1）%>%
选择（-value）
#hhid pid pname vid vname amt LOS
#           
#15668 93660 AB 462962 ABCD 20000 me
#2559593660汞柱608757 KJUI 20000 me
#34724100960 DC 772284后14000 ls
#4 4756 100960 DC 772284后14000 ls
#54856100960直流772284后14000低
#6 4730 100960 DC 293475 KAJA 14000 ls
#7 4757 100960 DC 293475 KAJA 14000 lg
#8 6320 100962 BA 656323 ADAR 27000 me
#9 4758 100960 DC 293475 KAJA 14000 ls
#10631100962 BA 656323阿达尔27000 lg
#116311100962BA81533SANG27000LO
#12547793814 BH 465612 NARI 20000 me
#13 6322 100962 BA 656323 ADAR 23000 ls
您可以执行以下操作：

df$LOS <- c("me", "ls", "lg", "lo")[apply(df[, c("me", "ls", "lg", "lo")]==1, 1, which)]
# > df
#    hhid    pid pname    vid vname me ls lg lo   amt LOS
# 1  5668  93660    AB 462962  ABCD  1  0  0  0 20000  me
# 2  5595  93660    HG 608757  KJUI  1  0  0  0 20000  me
# 3  4724 100960    DC 772284  HIND  0  1  0  0 14000  ls
# 4  4756 100960    DC 772284  HIND  0  1  0  0 14000  ls
# 5  4856 100960    DC 772284  HIND  0  0  0  1 14000  lo
# 6  4730 100960    DC 293475  KAJA  0  1  0  0 14000  ls
# 7  4757 100960    DC 293475  KAJA  0  0  1  0 14000  lg
# 8  6320 100962    BA 656323  ADAR  1  0  0  0 27000  me
# 9  4758 100960    DC 293475  KAJA  0  1  0  0 14000  ls
# 10 6319 100962    BA 656323  ADAR  0  0  1  0 27000  lg
# 11 6311 100962    BA  81533  SANG  0  0  0  1 27000  lo
# 12 5477  93814    BH 465612  NARI  1  0  0  0 20000  me
# 13 6322 100962    BA 656323  ADAR  0  1  0  0 23000  ls

df$LOS-df
#hhid pid pname vid vname me ls lg lo amt LOS
#15668 93660 AB 462962 ABCD 10020000 me
#259593660汞柱608757 KJUI 10020000ME
#3 4724 100960 DC 772284后01 0 14000 ls
#4756 100960 DC 772284后01 0 14000 ls
#54856100960直流772284后01 14000本地
#6 4730 100960 DC 293475卡贾01 0 14000 ls
#7 4757 100960 DC 293475卡加0 1 0 14000 lg
#8 6320 100962 BA 656323 ADAR 1 0 0 27000 me
#94758 100960 DC 293475卡贾01 0 14000 ls
#10631100962 BA 656323 ADAR 0 0 1 0 27000 lg
#116311100962BA81533SANG0127000LO
#12547793814伯克希尔哈撒韦465612纳里10020000ME
#13 6322 100962 BA 656323 ADAR 01 0 23000 ls

或（定义列名称向量）

cols您可以执行以下操作：

df$LOS <- c("me", "ls", "lg", "lo")[apply(df[, c("me", "ls", "lg", "lo")]==1, 1, which)]
# > df
#    hhid    pid pname    vid vname me ls lg lo   amt LOS
# 1  5668  93660    AB 462962  ABCD  1  0  0  0 20000  me
# 2  5595  93660    HG 608757  KJUI  1  0  0  0 20000  me
# 3  4724 100960    DC 772284  HIND  0  1  0  0 14000  ls
# 4  4756 100960    DC 772284  HIND  0  1  0  0 14000  ls
# 5  4856 100960    DC 772284  HIND  0  0  0  1 14000  lo
# 6  4730 100960    DC 293475  KAJA  0  1  0  0 14000  ls
# 7  4757 100960    DC 293475  KAJA  0  0  1  0 14000  lg
# 8  6320 100962    BA 656323  ADAR  1  0  0  0 27000  me
# 9  4758 100960    DC 293475  KAJA  0  1  0  0 14000  ls
# 10 6319 100962    BA 656323  ADAR  0  0  1  0 27000  lg
# 11 6311 100962    BA  81533  SANG  0  0  0  1 27000  lo
# 12 5477  93814    BH 465612  NARI  1  0  0  0 20000  me
# 13 6322 100962    BA 656323  ADAR  0  1  0  0 23000  ls

df$LOS-df
#hhid pid pname vid vname me ls lg lo amt LOS
#15668 93660 AB 462962 ABCD 10020000 me
#259593660汞柱608757 KJUI 10020000ME
#3 4724 100960 DC 772284后01 0 14000 ls
#4756 100960 DC 772284后01 0 14000 ls
#54856100960直流772284后01 14000本地
#6 4730 100960 DC 293475卡贾01 0 14000 ls
#7 4757 100960 DC 293475卡加0 1 0 14000 lg
#8 6320 100962 BA 656323 ADAR 1 0 0 27000 me
#94758 100960 DC 293475卡贾01 0 14000 ls
#10631100962 BA 656323 ADAR 0 0 1 0 27000 lg
#116311100962BA81533SANG0127000LO
#12547793814伯克希尔哈撒韦465612纳里10020000ME
#13 6322 100962 BA 656323 ADAR 01 0 23000 ls

或（定义列名称向量）

colsBase R溶液：

df <- transform(df[!(names(df) %in% c("me", "ls", "lg", "lo"))], 

                name = names(df)[(names(df) %in% c("me", "ls", "lg", "lo"))][apply(df[(names(df) %in% c("me", "ls", "lg", "lo"))], 1, which.max)])

dfBase R解决方案：

df <- transform(df[!(names(df) %in% c("me", "ls", "lg", "lo"))], 

                name = names(df)[(names(df) %in% c("me", "ls", "lg", "lo"))][apply(df[(names(df) %in% c("me", "ls", "lg", "lo"))], 1, which.max)])

df