Create a counter column in R Data.Table, with a condition on another column

Question

I am trying to count the years a product has been on offer since recorded history.

In the example below, I'd like to start the counter at the first non-zero sales recorded.

Note that I may have zero recorded sales in the next few years, but the counter should count the years as valid.

I experimented with rleid , but cannot figure out a way to differentiate the initial 0s vs. within-lifespan 0s.

tt <- data.table(YEAR=2007:2018,
                 SALES=c(0,0,0,2,3,5,1,0,9,0,3,4),
                 YEARS_IN=c(0,0,0,1,2,3,4,5,6,7,8,9))

> tt
    YEAR SALES YEARS_IN
 1: 2007     0        0
 2: 2008     0        0
 3: 2009     0        0
 4: 2010     2        1
 5: 2011     3        2
 6: 2012     5        3
 7: 2013     1        4
 8: 2014     0        5
 9: 2015     9        6
10: 2016     0        7
11: 2017     3        8
12: 2018     4        9

Answer 1

Here's a one line solution using base R -

tt <- data.table::data.table(YEAR=2007:2018,
                 SALES=c(0,0,0,2,3,5,1,0,9,0,3,4),
                 YEARS_IN=c(0,0,0,1,2,3,4,5,6,7,8,9))

tt$Calc_Years <- cumsum(cumsum(tt$SALES) > 0)
tt
    YEAR SALES YEARS_IN Calc_Years
 1: 2007     0        0          0
 2: 2008     0        0          0
 3: 2009     0        0          0
 4: 2010     2        1          1
 5: 2011     3        2          2
 6: 2012     5        3          3
 7: 2013     1        4          4
 8: 2014     0        5          5
 9: 2015     9        6          6
10: 2016     0        7          7
11: 2017     3        8          8
12: 2018     4        9          9

Thanks to Simon, here's a data.table version -

tt[ , Calc_Years := cumsum(cumsum(SALES) > 0)]

Answer 2

library(data.table)
tt <- data.table(YEAR=2007:2018,
                 SALES=c(0,0,0,2,3,5,1,0,9,0,3,4),
                 YEARS_IN=c(0,0,0,1,2,3,4,5,6,7,8,9))
temp <- min(which(tt[,SALES]!=0))
tt[-c(1:(temp-1)),Comp:=.I]
tt[c(1:(temp-1)),Comp:=0]
tt

    YEAR SALES YEARS_IN Comp
 1: 2007     0        0    0
 2: 2008     0        0    0
 3: 2009     0        0    0
 4: 2010     2        1    1
 5: 2011     3        2    2
 6: 2012     5        3    3
 7: 2013     1        4    4
 8: 2014     0        5    5
 9: 2015     9        6    6
10: 2016     0        7    7
11: 2017     3        8    8
12: 2018     4        9    9