0

I have a series of animal observations linked with the time of the day. I need to group them in the following way for successive analyses; all the observations that are less than 10 minutes apart from the previous are in the same group. When an observation is more than ten minutes apart from the previous it starts a new group.

For this I've done a for loop with an if statement

wilde16 <- structure(list(Species = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "wildebeest", class = "factor"), 
    Total_Ind = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 1L, 1L, 
    1L, 1L, 1L, 1L, 1L), DateTime = structure(c(1464449354, 1464449370, 
    1464449393, 1464449402, 1464449876, 1464449954, 1464450303, 
    1464450624, 1464450640, 1464450656, 1464450953, 1464450969, 
    1464450985, 1464451132, 1464451148, 1464451165), class = c("POSIXct", 
    "POSIXt"), tzone = "Africa/Dar_es_Salaam")), row.names = c(99L, 
100L, 101L, 102L, 103L, 104L, 105L, 106L, 107L, 110L, 128L, 129L, 
132L, 142L, 144L, 146L), class = "data.frame")    

class(wilde16$DateTime)
[1] "POSIXct" "POSIXt" 

wilde16$DateTime[223]
[1] "2016-05-30 09:54:54 EAT"

z <- 0

for(i in 1:length(wilde16$DateTime)){
if(wilde16$DateTime[i+1]-wilde16$DateTime[i]<600){
    wilde16$Group[i] <- z
  } 
  else{ 
    z <- z + 1
    wilde16$Group[i] <- z
    }
 }

Yet when I run it returns the error message

"Error in if (wilde16$DateTime[i + 1] - wilde16$DateTime[i] < 600) { : missing value where TRUE/FALSE needed"

Even though if I try the lines

i <- 1
(wilde16$DateTime[i + 1] - wilde16$DateTime[i] < 600)

it returns

[1] FALSE

What's wrong?

Tommaso Saccà
  • 15
  • 4
  • 2
    What do you think happens to `wilde16$DateTime[i+1]` when `i` is `length(wilde16$DateTime)` ? What do you want to happen? – Henry Jun 09 '19 at 15:58
  • 2
    Please [edit your post to improve reproducibility](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example). There might certainly be no need to write a loop. – NelsonGon Jun 09 '19 at 16:16
  • Thanks for the link @Nelson, now my question should be better – Tommaso Saccà Jun 11 '19 at 14:42
  • @Henry, following the line of the suggestion I changed the 'i' from '1 : length(wilde16$DateTime)' to '1 : (length(wilde16@DateTime) - 1)'. Now it stil doesn't provide the grouping that I intended but the loop runs without errors. Luckily i94pxoe code works as a charm. – Tommaso Saccà Jun 11 '19 at 14:56

1 Answers1

1

As @Henry observed, you have a index issue.

If you are trying to find intervals on your vector, you can probably avoid the for loop by using the function findInterval(). 

interval <- seq.POSIXt(from = min(wilde16$DateTime), to = max(wilde16$DateTime), by = "10 min")
z <- findInterval(x = wilde16$DateTime, vec = interval)
NelsonGon
  • 13,015
  • 7
  • 27
  • 57
i94pxoe
  • 577
  • 3
  • 11