用R编程中的条件循环

Question

I would like to compare the previous row value whether it is same as the current one (for more than 1 variables and also using list of values). In this case how do I perform write code. I read 'apply' functions can be used.

I searched this topic here before posting this question found somewhat similar but unable to find the exact one. I'm quite new to R.

Here is my sample table: (Flag needs to be done based on conditions)

Ticket No   V1  V2  Flag
Tkt10256    1   X   0
Tkt10257    1   aa  0
Tkt10257    2   bb  1
Tkt10257    3   x   0
Tkt10260    1   cc  0
Tkt10260    2   aa  1
Tkt10262    3   bb  0

I have to Flag based on the below conditions (if all the conditions are satisfied then mark as 1)

1. Variable 2 should be the following one of 4 names (aa, bb, cc, dd)
2. Variable 1 should be the different from previous row
3. Ticket number has to be the same as previous row

Thanks in advance for the help !

Answer 1

An approach without looping:

indx1 <- with(df, V2 %in% paste0(letters[1:4], letters[1:4]) )
indx2 <- with(df, c(TRUE,V1[-1]!=V1[-length(V1)]))
indx3 <- with(df, c(FALSE,Ticket.No[-1]==Ticket.No[-nrow(df)]))

df$Flag <- (indx1 & indx2 & indx3)+0
df$Flag
#[1] 0 0 1 0 0 1 0

data

df <- structure(list(Ticket.No = c("Tkt10256", "Tkt10257", "Tkt10257", 
"Tkt10257", "Tkt10260", "Tkt10260", "Tkt10262"), V1 = c(1L, 1L, 
2L, 3L, 1L, 2L, 3L), V2 = c("X", "aa", "bb", "x", "cc", "aa", 
"bb"), Flag = c(0L, 0L, 1L, 1L, 0L, 1L, 0L)), .Names = c("Ticket.No", 
"V1", "V2", "Flag"), class = "data.frame", row.names = c(NA, 
-7L))

Answer 2

One more:

Check this on your larger data. I'm not exactly sure if duplicated is the right function to use there. If the numbers in the TicketNo column are increasing (ie the Xs in TktXXXXX), then it should work fine.

> dat2 <- dat[dat$V2 %in% c("aa", "bb", "cc", "dd"),]
> rn <- rownames(dat2)[duplicated(dat2[[1]]) & !c(FALSE, diff(dat2[[2]]) == 0)]
> dat$Flag <- (rownames(dat) %in% rn)+0
> dat
#   TicketNo V1 V2 Flag
# 1 Tkt10256  1  X    0
# 2 Tkt10257  1 aa    0
# 3 Tkt10257  2 bb    1
# 4 Tkt10257  3  x    0
# 5 Tkt10260  1 cc    0
# 6 Tkt10260  2 aa    1
# 7 Tkt10262  3 bb    0

Answer 3

A variation on @Akrun's answer:

with(df, 
  V2 %in% c("aa","bb","cc","dd") &  
  c(FALSE,diff(V1) != 0) &
  c(FALSE,head(Ticket.No, -1)) == Ticket.No
) + 0

#[1] 0 0 1 0 0 1 0

Answer 4

Try:

for(i in 2:nrow(ddf)){
   ddf$Flag[i] = ifelse(  ddf$V2[i] %in% c('aa', 'bb', 'cc', 'dd') 
           && ddf$V1[i] != ddf$V1[(i-1)] 
           &&  ddf$TicketNo[i] == ddf$TicketNo[(i-1)]
         ,1,0)
 }
ddf
  TicketNo V1 V2 Flag
1 Tkt10256  1  X    0
2 Tkt10257  1 aa    0
3 Tkt10257  2 bb    1
4 Tkt10257  3  x    0
5 Tkt10260  1 cc    0
6 Tkt10260  2 aa    1
7 Tkt10262  3 bb    0