pkg/R/getData.R

   1 #' @title Acquire data in a clean format
   2 #'
   3 #' @description Take in input data frames and/or files containing raw data, and timezones, and
   4 #'   output a Data object, roughly corresponding to a list where each cell contains all value
   5 #'   for one day (see \code{?Data}).
   6 #'
   7 #' @param ts_data Time-series, as a data frame (DB style: 2 columns, first is date/time,
   8 #'   second is value) or a CSV file
   9 #' @param exo_data Exogenous variables, as a data frame or a CSV file; first comlumn is dates,
  10 #'   next block are measurements for the day, and final block are exogenous forecasts
  11 #' @param input_tz Timezone in the input files ("GMT" or e.g. "Europe/Paris")
  12 #' @param date_format How date/time are stored (e.g. year/month/day hour:minutes;
  13 #'   see \code{strptime})
  14 #' @param working_tz Timezone to work with ("GMT" or e.g. "Europe/Paris")
  15 #' @param predict_at When does the prediction take place ? Integer, in hours. Default: 0
  16 #' @param limit Number of days to extract (default: Inf, for "all")
  17 #'
  18 #' @return An object of class Data
  19 #'
  20 #' @examples
  21 #' ts_data = read.csv(system.file("extdata","pm10_mesures_H_loc.csv",package="talweg"))
  22 #' exo_data = read.csv(system.file("extdata","meteo_extra_noNAs.csv",package="talweg"))
  23 #' data = getData(ts_data, exo_data, limit=120)
  24 #' @export
  25 getData = function(ts_data, exo_data, input_tz="GMT", date_format="%d/%m/%Y %H:%M",
  26     working_tz="GMT", predict_at=0, limit=Inf)
  27 {
  28     # Sanity checks (not full, but sufficient at this stage)
  29     if (!is.character(input_tz) || !is.character(working_tz))
  30         stop("Bad timezone (see ?timezone)")
  31     input_tz = input_tz[1]
  32     working_tz = working_tz[1]
  33     if ( (!is.data.frame(ts_data) && !is.character(ts_data)) ||
  34             (!is.data.frame(exo_data) && !is.character(exo_data)) )
  35         stop("Bad time-series / exogenous input (data frame or CSV file)")
  36     if (is.character(ts_data))
  37         ts_data = ts_data[1]
  38     if (is.character(exo_data))
  39         exo_data = exo_data[1]
  40     predict_at = as.integer(predict_at)[1]
  41     if (predict_at<0 || predict_at>23)
  42         stop("Bad predict_at (0-23)")
  43     if (!is.character(date_format))
  44         stop("Bad date_format (character)")
  45     date_format = date_format[1]
  46
  47     ts_df =
  48         if (is.character(ts_data))
  49             read.csv(ts_data)
  50         else
  51             ts_data
  52     # Convert to the desired timezone (usually "GMT" or "Europe/Paris")
  53     formatted_dates_POSIXlt = strptime(as.character(ts_df[,1]), date_format, tz=input_tz)
  54     ts_df[,1] = format(as.POSIXct(formatted_dates_POSIXlt, tz=input_tz), tz=working_tz, usetz=TRUE)
  55
  56     exo_df =
  57         if (is.character(exo_data))
  58             read.csv(exo_data)
  59         else
  60             exo_data
  61     # Times in exogenous variables file are ignored: no conversions required
  62
  63     line = 1 #index in PM10 file (24 lines for 1 cell)
  64     nb_lines = nrow(ts_df)
  65     nb_exos = ( ncol(exo_df) - 1 ) / 2
  66     data = Data$new()
  67     i = 1 #index of a cell in data
  68     while (line <= nb_lines)
  69     {
  70         time = c()
  71         serie = c()
  72         repeat
  73         {
  74             {
  75                 time = c(time, ts_df[line,1])
  76                 serie = c(serie, ts_df[line,2])
  77                 line = line + 1
  78             };
  79             if (line >= nb_lines + 1 || as.POSIXlt(ts_df[line-1,1],tz=working_tz)$hour == predict_at)
  80                 break
  81         }
  82
  83         exo = as.data.frame( exo_df[i,2:(1+nb_exos)] )
  84         exo_hat = as.data.frame( exo_df[i,(1+nb_exos+1):(1+2*nb_exos)] )
  85         level = mean(serie, na.rm=TRUE)
  86         centered_serie = serie - level
  87         data$append(time, centered_serie, level, exo, exo_hat)
  88         if (i >= limit)
  89             break
  90         i = i + 1
  91     }
  92     if (length(data$getCenteredSerie(1)) < length(data$getCenteredSerie(2)))
  93         data$removeFirst()
  94     if (length(data$getCenteredSerie(data$getSize()))
  95         < length(data$getCenteredSerie(data$getSize()-1)))
  96     {
  97         data$removeLast()
  98     }
  99     data
 100 }