cleaning - fix getSimilarDaysIndices ; to finish
[talweg.git] / 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 }