Gettting cumulative sum of previous values except for the first value

Question

I am running a code in R, the sample of which is as below with a small dataset -

library(plyr)
Ex<-structure(list(X1 = c(-36.8598, -37.1726, -36.4343, -36.8644, 
                      -37.0599, -34.8818, -31.9907, -37.8304, 
                      -34.3367, -31.2984, -33.5731), 
               X2 = c(64.26, 63.085, 66.36, 61.08, 61.57, 65.04, 72.69, 63.83, 
                      67.555, 76.06, 68.61), 
               Y1 = c(493.81544, 493.81544, 494.54173, 
                      494.61364, 494.61381, 494.38717, 494.64122, 493.73265,             494.04246, 
                      494.92989, 494.98384), 
               Y2 = c(489.704166, 489.704166, 490.710962, 
                      490.653212, 490.710612, 489.822928, 
                      488.160904, 489.747776, 490.600579, 
                      488.946738, 490.398958), 
               Y3 = c(19L, 19L, 19L, 23L, 30L,43L,43L,2L, 58L, 47L, 61L),
               date = c("2013-06-01","2013-06-02","2013-06-03","2013-06-04",
                        "2013-06-05","2013-06-06","2013-06-07","2013-06-08",
                        "2013-06-09","2013-06-10","2013-06-11")), 
          .Names = c("X1", "X2", "Y1", "Y2", "Y3", "date"), 
          row.names = c(NA, 11L), class = "data.frame")

Ex <- arrange(Ex, Y3)

Ex$Dup <- as.numeric(duplicated(Y3))
Ex$Dup_rev <- as.numeric(duplicated(Y3,fromLast=TRUE))

##Testing If Else
attach(Ex)
Ex$X5 <- 0
for(i in 1:length(Y3))
{
  if (Ex$Dup[i]==0 & Ex$Dup_rev[i]==0)
  {
    Ex$X5[i]=Y2[i]
  } else if(Ex$Dup[i]==0)
  {
    Ex$X5[i]=Y2[i]
  }else 
    {Ex$X5[i]=Y2[i] + X5[i-1]}
}

What this does is unless the value of column Y3 is the first time it appears in the data set, for every row of Y3, we need to create a column X5 that is the cumulative sum of previous Y2. Since my data is huge(~110k rows of data), this code is taking a hell lot of time to execute. Is there any simpler way to execute the same code?

X1  X2  Y1  Y2  Y3  date    Dup Dup_rev X5
 1  -37.8304    63.830  493.7326    489.7478    2   2013-06-08  0   0   489.7478
 2  -36.8598    64.260  493.8154    489.7042    19  2013-06-01  0   1   489.7042
 3  -37.1726    63.085  493.8154    489.7042    19  2013-06-02  1   1   1469.1125
 4  -36.4343    66.360  494.5417    490.7110    19  2013-06-03  1   0   1470.1193
 5  -36.8644    61.080  494.6136    490.6532    23  2013-06-04  0   0   490.6532

Answer 1

Here's a solution with dplyr. dplyr is the next iteration of plyr and is very fast.

library(dplyr)
Ex %.% group_by(Y3) %.% mutate(X5 = cumsum(Y2))
#> Source: local data frame [11 x 7]
#> Groups: Y3
#> 
#>          X1     X2       Y1       Y2 Y3       date        X5
#> 1  -36.8598 64.260 493.8154 489.7042 19 2013-06-01  489.7042
#> 2  -37.1726 63.085 493.8154 489.7042 19 2013-06-02  979.4083
#> 3  -36.4343 66.360 494.5417 490.7110 19 2013-06-03 1470.1193
#> 4  -36.8644 61.080 494.6136 490.6532 23 2013-06-04  490.6532
#> 5  -37.0599 61.570 494.6138 490.7106 30 2013-06-05  490.7106
#> 6  -34.8818 65.040 494.3872 489.8229 43 2013-06-06  489.8229
#> 7  -31.9907 72.690 494.6412 488.1609 43 2013-06-07  977.9838
#> 8  -37.8304 63.830 493.7326 489.7478  2 2013-06-08  489.7478
#> 9  -34.3367 67.555 494.0425 490.6006 58 2013-06-09  490.6006
#> 10 -31.2984 76.060 494.9299 488.9467 47 2013-06-10  488.9467
#> 11 -33.5731 68.610 494.9838 490.3990 61 2013-06-11  490.3990

Answer 2

Here is a solution using data.table, which is very fast for this type of analysis were you split by a "factor" (in this case Y3):

library(data.table)
DT <- data.table(Ex)[, X5:=cumsum(Y2), by=Y3]
DT
#           X1     X2       Y1       Y2 Y3       date        X5
#  1: -37.8304 63.830 493.7326 489.7478  2 2013-06-08  489.7478
#  2: -36.8598 64.260 493.8154 489.7042 19 2013-06-01  489.7042
#  3: -37.1726 63.085 493.8154 489.7042 19 2013-06-02  979.4083
#  4: -36.4343 66.360 494.5417 490.7110 19 2013-06-03 1470.1193
#  5: -36.8644 61.080 494.6136 490.6532 23 2013-06-04  490.6532
#  6: -37.0599 61.570 494.6138 490.7106 30 2013-06-05  490.7106
#  7: -34.8818 65.040 494.3872 489.8229 43 2013-06-06  489.8229
#  8: -31.9907 72.690 494.6412 488.1609 43 2013-06-07  977.9838
#  9: -31.2984 76.060 494.9299 488.9467 47 2013-06-10  488.9467
# 10: -34.3367 67.555 494.0425 490.6006 58 2013-06-09  490.6006
# 11: -33.5731 68.610 494.9838 490.3990 61 2013-06-11  490.3990    

Note though that like Jake, I don't understand how you get 1469 for the third row instead of 979.4083. Also, I just ran your code and got the same answer as I did, so I'm guessing there is a typo in your sample result, or perhaps the data changed?