在我的data.frame中，我在一行的SUBJECT上有三列。我想要一个额外的列，每行都有一个唯一的主题。首先，我的数据如何：
DATE <- c("1","2","3","4","5","6","7","1","2","3","4","5","6","7")
COMP <- c("A", "A", "A", "A", "A", "A", "A", "B", "B", "B", "B", "B", "B", "B")
RET <- c(-2.0,1.1,3,1.4,-0.2, 0.6, 0.1, -0.21, -1.2, 0.9, 0.3, -0.1,0.3,-0.12)
CLASS <- c("positive", "negative", "aneutral", "positive", "positive", "negative", "aneutral", "positive", "negative", "negative", "positive", "aneutral", "aneutral", "aneutral")
SUBJECT.1 <- c("LITIGATION","LAYOFF","POLLUTION","CHEMICAL DISASTER","PRESS RELEASE","PEOPLE","EMISSIONS","ENERGY","WASTE MANAGEMENT","EMPLOYEES","MANAGEMENT","PRESS RELEASE","HOTELS","POLLUTION")
SUBJECT.2 <- c("POLLUTION","EMPLOYEES","NUCLEAR","FUELS","STOCK OPTION PLAN","EXECUTIVES","CO2","SOLAR","POLLUTION","EXECUTIVES","PRESS RELEASE","CELEBRITIES","CELEBRITIES","LITIGATION")
SUBJECT.3 <- c("ENVIRONMENT","JOB REDUCTIONS","POWER PLANTS","POLLUTION","EMPLOYEES","FRAUD","CLIMATE CHANGE","SUSTAINABILITY","HAZARDOUS WASTE","BONUS PAY","LITIGATION","EMISSIONS","SCANDALS","SCANDALS")
CONTROLVAR <- c("11","13","13","14","13","14","12","11","13","13","14","13","14","12")
mydf <- data.frame(DATE, COMP, RET, CLASS, SUBJECT.1, SUBJECT.2, SUBJECT.3, CONTROLVAR, stringsAsFactors=F)
mydf
# DATE COMP RET CLASS SUBJECT.1 SUBJECT.2 SUBJECT.3 CONTROLVAR
# 1 1 A -2.00 positive LITIGATION POLLUTION ENVIRONMENT 11
# 2 2 A 1.10 negative LAYOFF EMPLOYEES JOB REDUCTIONS 13
# 3 3 A 3.00 aneutral POLLUTION NUCLEAR POWER PLANTS 13
# 4 4 A 1.40 positive CHEMICAL DISASTER FUELS POLLUTION 14
# 5 5 A -0.20 positive PRESS RELEASE STOCK OPTION PLAN EMPLOYEES 13
# 6 6 A 0.60 negative PEOPLE EXECUTIVES FRAUD 14
# 7 7 A 0.10 aneutral EMISSIONS CO2 CLIMATE CHANGE 12
# 8 1 B -0.21 positive ENERGY SOLAR SUSTAINABILITY 11
# 9 2 B -1.20 negative WASTE MANAGEMENT POLLUTION HAZARDOUS WASTE 13
# 10 3 B 0.90 negative EMPLOYEES EXECUTIVES BONUS PAY 13
# 11 4 B 0.30 positive MANAGEMENT PRESS RELEASE LITIGATION 14
# 12 5 B -0.10 aneutral PRESS RELEASE CELEBRITIES EMISSIONS 13
# 13 6 B 0.30 aneutral HOTELS CELEBRITIES SCANDALS 14
# 14 7 B -0.12 aneutral POLLUTION LITIGATION SCANDALS 12
由于我想将主题作为虚拟变量（应该是独占的）包含在以后的回归中，我想要一个单独的列SUBJECT，每行有一个唯一的主题。我想专注于主题诉讼，污染和LAYOFF。
我想从左到右检查每个SUBJECT栏中的LITIGATION，POLLUTION和LAYOFF。
如果在第一栏中有三个主题中的一个，即诉讼，污染或LAYOFF，则采取此主题。如果第一列中有不同的主题，我会检查第二列，依此类推。如果三个主题栏中没有一个包含LITIGATION，POLLUTION或LAYOFF，则该主题应称为OTHER。 此外，一些主题应该分组。在这个例子中，应该将EMISSIONS视为POLLUTION。
输出应如下所示：
# DATE COMP RET CLASS SUBJECT.1 SUBJECT.2 SUBJECT.3 SUBJECT CONTROLVAR
# 1 1 A -2.00 positive LITIGATION POLLUTION ENVIRONMENT LITIGATION 11
# 2 2 A 1.10 negative LAYOFF EMPLOYEES JOB REDUCTIONS LAYOFF 13
# 3 3 A 3.00 aneutral POLLUTION NUCLEAR POWER PLANTS POLLUTION 13
# 4 4 A 1.40 positive CHEMICAL DISASTER FUELS POLLUTION POLLUTION 14
# 5 5 A -0.20 positive PRESS RELEASE STOCK OPTION PLAN EMPLOYEES OTHER 13
# 6 6 A 0.60 negative PEOPLE EXECUTIVES FRAUD OTHER 14
# 7 7 A 0.10 aneutral EMISSIONS CO2 CLIMATE CHANGE POLLUTION 12
# 8 1 B -0.21 positive ENERGY SOLAR SUSTAINABILITY OTHER 11
# 9 2 B -1.20 negative WASTE MANAGEMENT POLLUTION HAZARDOUS WASTE POLLUTION 13
# 10 3 B 0.90 negative EMPLOYEES EXECUTIVES BONUS PAY OTHER 13
# 11 4 B 0.30 positive MANAGEMENT PRESS RELEASE LITIGATION LITIGATION 14
# 12 5 B -0.10 aneutral PRESS RELEASE CELEBRITIES EMISSIONS POLLUTION 13
# 13 6 B 0.30 aneutral HOTELS CELEBRITIES SCANDALS OTHER 14
# 14 7 B -0.12 aneutral POLLUTION LITIGATION SCANDALS POLLUTION 12
谢谢！
答案 0 :(得分：2)
mydf$SUBJECT <- "OTHER"
sapply(c("SUBJECT.3", "SUBJECT.2", "SUBJECT.1"), function(x) mydf[mydf[, x] %in% c("LITIGATION", "POLLUTION", "LAYOFF", "EMISSIONS"), "SUBJECT"] <<- mydf[mydf[, x] %in% c("LITIGATION", "POLLUTION", "LAYOFF", "EMISSIONS"), x])
mydf$SUBJECT[mydf$SUBJECT == "EMISSIONS"] <- "POLLUTION"
答案 1 :(得分：0)
这里有几行解决方案，只需将3 ifelse()链接在一起。
important <- c('LITIGATION','POLLUTION','LAYOFF','EMISSIONS')
ifelse( mydf$SUBJECT.1 %in% important, mydf$SUBJECT.1,
ifelse( mydf$SUBJECT.2 %in% important, mydf$SUBJECT.2,
ifelse( mydf$SUBJECT.3 %in% important, mydf$SUBJECT.3,'OTHER')))
mydf$SUBJECT[mydf$SUBJECT=='EMISSIONS'] <- 'POLUTION'