A function to find continuous sequences from a vector.

find.sequences.R

#
#  "`-''-/").___..--''"`-._
# (`6_ 6  )   `-.  (     ).`-.__.`)   WE ARE ...
# (_Y_.)'  ._   )  `._ `. ``-..-'    PENN STATE!
#   _ ..`--'_..-_/  /--'_.' ,'
# (il),-''  (li),'  ((!.-'
#
#
# Author: Weiming Hu (weiming@psu.edu) and Martina Calovi (mxc895@psu.edu)
#         Geoinformatics and Earth Observation Laboratory (http://geolab.psu.edu)
#         Department of Geography and Institute for CyberScience
#         The Pennsylvania State University
#

#' find.sequences
#' 
#' find.sequences finds the start and end indices for
#' continuous sequences that have a length larger than
#' `min.len`. The values for the sequences should all
#' be larger than the threshold.
#' 
#' @param v A numeric vector
#' @param min.len The minimum length of a sequence to
#' be extracted.
#' @param threshold The minimum value for the sequence.
#' 
#' @return A data frame with the start and end indices.
#' 
#' @examples 
#' v <- sin(seq(1, 20, length.out = 200))
#' v[sample(length(v), floor(length(v) * 0.2))] <- NA
#' 
#' plot(v, type = 'b')
#' min.len <- 5
#' threshold <- 0.5
#' df <- find.sequences(v, 5, 0)
#' abline(v = df$Start.ID, col = 'green')
#' abline(v = df$End.ID, col = 'red')
#' 
#' @md
#' @export
find.sequences <- function(v, min.len, threshold) {
  
  start.id <- c()
  end.id <- c()
  
  continous.day.count <- 0
  record <- F
  
  for (i in 1:length(v)) {
    
    if (is.na(v[i])) {
      
      if (record) {
        record <- F
        
        if (continous.day.count >= min.len) {
          end.id <- c(end.id, i - 1)
        } else {
          start.id <- start.id[-length(start.id)]
        }
      }
      
      continous.day.count <- 0
      
    } else {
      if (v[i] > threshold) {
        continous.day.count <- continous.day.count + 1
        
        if (!record) {
          record <- T
          start.id <- c(start.id, i)
        }
        
      } else {
        
        if (record) {
          record <- F
          
          if (continous.day.count >= min.len) {
            end.id <- c(end.id, i - 1)
          } else {
            start.id <- start.id[-length(start.id)]
          }
          
        }
        
        continous.day.count <- 0
      }
    }
  }
  
  if (record) {
    if (continous.day.count >= min.len) {
      end.id <- c(end.id, i)
    } else {
      start.id <- start.id[-length(start.id)]
    }
  }
  
  return(data.frame(Start.ID = start.id,
                    End.ID = end.id))
}