How to find the indices where there are n consecutive zeroes in a row

Question

Suppose I have this data:

  x = c(14,14, 6,  7 ,14 , 0 ,0  ,0 , 0,  0,  0 , 0 , 0,  0 , 0 , 0 , 0,  9  ,1 , 3  ,8  ,9 ,15,  9 , 8, 13,  8,  4 , 6 , 7 ,10 ,13,  3,
 0 , 0 , 0 , 0 , 0 , 0,  0,  0 , 0 , 0 , 0,  0,  0,  0,  0  ,0,  0 , 0 , 0,  0,  0,  0,  0 , 0,  0, 4 , 7  ,4,  5 ,16 , 5  ,5 , 9 , 4  ,4,  9 , 8,  2,  0  ,0  ,0  ,0  ,0,  0,  0,  0  ,0 , 0,  0,  0,  0,  0,  0,  0,  0,0)

x
 [1] 14 14  6  7 14  0  0  0  0  0  0  0  0  0  0  0  0  9  1  3  8  9 15  9  8
[26] 13  8  4  6  7 10 13  3  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0
[51]  0  0  0  0  0  0  0  0  4  7  4  5 16  5  5  9  4  4  9  8  2  0  0  0  0
[76]  0  0  0  0  0  0  0  0  0  0  0  0  0  0  

I want to recover the indices beginning where there are more than 3 zeroes in a row and terminating with the last 0 before a nonzero.

For example,

I would get

6, 17 for the first rash of zeroes, etc.

Answer 1

If x happens to be a column of a data.table you can do

library(data.table)
dt <- data.table(x = x)

dt[, if(.N > 3 & all(x == 0)) .(starts = first(.I), ends = last(.I))
   , by = rleid(x)]

#    rleid starts ends
# 1:     5      6   17
# 2:    22     34   58
# 3:    34     72   89

Explanation:

• rleid(x) gives an ID (integer) for each element in x indicating which "run" the element is a member of, where "run" means a sequence of adjacent equal values.

• dt[, <code>, by = rle(x)] partitions dt according to rleid(x) and computes <code> for each subset of dt's rows. The results are stacked together in a single data.table.

• .N is the number of elements in the given subset

• .I is the vector of row numbers corresponding to the subset

• first and last give the first and last element of a vector

• .(<stuff>) is the same as list(<stuff>)

  The rleid function, by grouping within the brackets, .N and .I symbols, first and last functions are part of the data.table package.

Answer 2

By using dplyr , get the diff then if the diff not equal to 0 , they are not belong to same group , after cumsum we get the grouid

library(dplyr)
df=data.frame('x'=x,rownumber=seq(length(x)))
df$Groupid=cumsum(c(0,diff(df$x==0))!=0)
df%>%group_by(Groupid)%>%summarize(start=first(rownumber),end=last(rownumber),number=first(x),size=n())%>%filter(number==0&size>=3)
# A tibble: 3 x 5
  Groupid start   end number  size
    <int> <int> <int>  <dbl> <int>
1       1     6    17      0    12
2       3    34    58      0    25
3       5    72    89      0    18