R/flat.R

In INQC: Quality Control of Climatological Daily Time Series

flat<-function(y,maxseq,exclude=NULL){

  #' Flat sequences
  #' @description This function detects consecutive equal values (e.g., 15.1, 15.1, 15.1, 15.1...) in a data time series.
  #' Also can be used to detect consecutive equal decimal part of the values (e.g., 15.1, 12.1, 13.1, 10.1 ...)
  #' @param y data vector
  #' @param maxseq the maximum number of contiguous repetitions of a value (e.g., if 3, sequences of 4 will be flagged)
  #' @param exclude values to be excluded. This is useful for variables where a single value is expected to repeat many times, e.g. 0.0 in precipitation.
  #' @return  list of positions which do not pass this QC test. If all positions pass the test, returns NULL
  #' @examples
  #' y<-rnorm(100)
  #' y[10:20]<-10
  #' flat(y,5)
  #'
  #' #Extract the ECA&D data file from the example data folder
  #' path2inptfl<-system.file("extdata", "TX_SOUID132734.txt", package = "INQC")
  #' #Read the data file
  #' y<-readecad(input=path2inptfl,missing= -9999)[,4]
  #' #Find all consecutive (with a length > 5 elements) equal values in the time series
  #' flat(y,5)
  #'
  #' #Introduce the duplicated dates
  #' y[6:12]<-10
  #' #Find all consecutive (with a length > 5 elements) equal values in the time series
  #' flat(y,5)
  #' @export

  target<-NULL
  mu<-rle(y)$lengths # sequences
  cumu<-cumsum(mu) # cumulative sum to find real position of a sequence
  longbad<-mu[which(mu>maxseq)] # length of each sequence
  # make resistant to nullbad
  if(length(longbad)!=0){
    bad<-cumu[which(mu>maxseq)] # last point of a sequence
    firstbad=(bad-longbad)+1 # first point of a sequence
    ene<-length(firstbad)
    for(i in 1:ene){
      if(i==1){target<-firstbad[i]:bad[i]}else{target<-c(target,firstbad[i]:bad[i])}
    }
    if(!is.null(exclude)){
      quitar<-which(y[target] == exclude)
      target<-target[-quitar]
    }
    return(target)
  }
}