pkg/R/utils.R

   1 #' dateIndexToInteger
   2 #'
   3 #' Transform a (potential) date index into an integer (relative to data beginning).
   4 #'
   5 #' @param index Date (or integer) index
   6 #' @param data Object of class Data, output of \code{getData()}
   7 #'
   8 #' @export
   9 dateIndexToInteger = function(index, data)
  10 {
  11     # Works on integers too: trust input
  12     if (is.numeric(index))
  13         index = as.integer(index)
  14     if (is.integer(index))
  15         return (index)
  16
  17     if (inherits(index, "Date") || is.character(index))
  18     {
  19         tryCatch(indexAsDate <- as.Date(index),
  20             error=function(e) stop("Unrecognized index format"))
  21         #TODO: tz arg to difftime ?
  22         integerIndex <- round( as.numeric(
  23             difftime(indexAsDate, as.Date(data$getTime(1)[1])) ) ) + 1
  24         if (integerIndex >= 1 && integerIndex <= data$getSize())
  25             return (integerIndex)
  26         stop("Date outside data range")
  27     }
  28     stop("Unrecognized index format")
  29 }
  30
  31 #' integerIndexToDate
  32 #'
  33 #' Transform an integer index (relative to data beginning) into a date index.
  34 #'
  35 #' @inheritParams dateIndexToInteger
  36 #'
  37 #' @export
  38 integerIndexToDate = function(index, data)
  39 {
  40     # Works on dates too: trust input
  41     if (is.character(index))
  42         index = as.Date(index)
  43     if (is(index,"Date"))
  44         return (index)
  45
  46     index = index[1]
  47     if (is.numeric(index))
  48         index = as.integer(index)
  49     if (!is.integer(index))
  50         stop("'index' should be a date or integer")
  51     as.Date( data$getTime(index)[1] )
  52 }
  53
  54 #' getSimilarDaysIndices
  55 #'
  56 #' Find similar days indices in the past; at least same type of day in the week:
  57 #' monday=tuesday=wednesday=thursday != friday != saturday != sunday.
  58 #'
  59 #' @param index Day index (numeric or date)
  60 #' @param data Reference dataset, object output of \code{getData}
  61 #' @param limit Maximum number of indices to return
  62 #' @param same_season Should the indices correspond to day in same season?
  63 #' @param days_in Optional set to intersect with results (NULL to discard)
  64 #'
  65 #' @export
  66 getSimilarDaysIndices = function(index, data, limit, same_season,
  67     days_in=NULL, operational=TRUE)
  68 {
  69     index = dateIndexToInteger(index, data)
  70
  71     # Look for similar days (optionally in same season)
  72     days = c()
  73     dt_ref = as.POSIXlt(data$getTime(index)[1]) #first date-time of current day
  74     day_ref = dt_ref$wday #1=monday, ..., 6=saturday, 0=sunday
  75     month_ref = as.POSIXlt(data$getTime(index)[1])$mon+1 #month in 1...12
  76     i = index - 1
  77     if (!operational)
  78         j = index + 1
  79     while (length(days) < min( limit, ifelse(is.null(days_in),Inf,length(days_in)) ))
  80     {
  81         if (i < 1 && j > data$getSize())
  82             break
  83         if (i >= 1)
  84         {
  85             dt = as.POSIXlt(data$getTime(i)[1])
  86             if ((is.null(days_in) || i %in% days_in) && .isSameDay(dt$wday, day_ref))
  87             {
  88                 if (!same_season || .isSameSeason(dt$mon+1, month_ref))
  89                     days = c(days, i)
  90             }
  91             i = i - 1
  92         }
  93         if (!operational && j <= data$getSize())
  94         {
  95             dt = as.POSIXlt(data$getTime(j)[1])
  96             if ((is.null(days_in) || j %in% days_in) && .isSameDay(dt$wday, day_ref))
  97             {
  98                 if (!same_season || .isSameSeason(dt$mon+1, month_ref))
  99                     days = c(days, j)
 100             }
 101             j = j + 1
 102         }
 103     }
 104     return ( days )
 105 }
 106
 107 # isSameSeason
 108 #
 109 # Check if two months fall in the same "season" (defined by estimated pollution rate).
 110 #
 111 # @param month Month index to test
 112 # @param month_ref Month to compare to
 113 #
 114 .isSameSeason = function(month, month_ref)
 115 {
 116     if (month_ref == 3) #TODO: same as Bruno (but weird)
 117         return (month %in% c(2,3,4,9,10))
 118     if (month_ref %in% c(11,12,1,2)) #~= mid-polluted
 119         return (month %in% c(11,12,1,2))
 120     if (month_ref %in% c(3,4,9,10)) #~= high-polluted
 121         return (month %in% c(3,4,9,10))
 122     return (month %in% c(5,6,7,8)) #~= non polluted
 123 }
 124
 125 # isSameDay
 126 #
 127 # Monday=Tuesday=Wednesday=Thursday ; Friday, Saturday, Sunday: specials.
 128 #
 129 # @param day Day index to test
 130 # @param day_ref Day index to compare to
 131 #
 132 .isSameDay = function(day, day_ref)
 133 {
 134     if (day_ref %in% 1:4)
 135         return (day %in% 1:4)
 136     return (day == day_ref)
 137 }
 138
 139 # getNoNA2
 140 #
 141 # Get indices in data of no-NA series preceded by no-NA, within [first,last] range.
 142 #
 143 # @inheritParams dateIndexToInteger
 144 # @param first First index (included)
 145 # @param last Last index (included)
 146 #
 147 .getNoNA2 = function(data, first, last)
 148 {
 149     (first:last)[ sapply(first:last, function(i)
 150         !any( is.na(data$getSerie(i-1)) | is.na(data$getSerie(i)) )
 151     ) ]
 152 }