pkg/R/F_Neighbors2.R

   1 #' @include Forecaster.R
   2 #'
   3 #' Neighbors2 Forecaster
   4 #'
   5 #' Predict tomorrow as a weighted combination of "futures of the past" days.
   6 #' Inherits \code{\link{Forecaster}}
   7 #'
   8 Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster",
   9     inherit = Forecaster,
  10
  11     public = list(
  12         predictShape = function(data, today, memory, horizon, ...)
  13         {
  14             # (re)initialize computed parameters
  15             private$.params <- list("weights"=NA, "indices"=NA, "window"=NA)
  16
  17             # Do not forecast on days with NAs (TODO: softer condition...)
  18             if (any(is.na(data$getCenteredSerie(today))))
  19                 return (NA)
  20
  21             # Determine indices of no-NAs days followed by no-NAs tomorrows
  22             fdays = getNoNA2(data, max(today-memory,1), today-1)
  23
  24             # Get optional args
  25             simtype = ifelse(hasArg("simtype"), list(...)$simtype, "mix") #or "endo", or "exo"
  26             kernel = ifelse(hasArg("kernel"), list(...)$kernel, "Gauss") #or "Epan"
  27             if (hasArg(h_window))
  28             {
  29                 return ( private$.predictShapeAux(data,
  30                     fdays, today, horizon, list(...)$h_window, kernel, simtype, TRUE) )
  31             }
  32
  33             # Indices of similar days for cross-validation; TODO: 20 = magic number
  34             cv_days = getSimilarDaysIndices(today, data, limit=20, same_season=FALSE, days_in=fdays)
  35
  36             # Function to optimize h : h |--> sum of prediction errors on last 45 "similar" days
  37             errorOnLastNdays = function(h, kernel, simtype)
  38             {
  39                 error = 0
  40                 nb_jours = 0
  41                 for (i in seq_along(cv_days))
  42                 {
  43                     # mix_strategy is never used here (simtype != "mix"), therefore left blank
  44                     prediction = private$.predictShapeAux(data,
  45                         fdays, cv_days[i], horizon, h, kernel, simtype, FALSE)
  46                     if (!is.na(prediction[1]))
  47                     {
  48                         nb_jours = nb_jours + 1
  49                         error = error +
  50                             mean((data$getCenteredSerie(cv_days[i]+1)[1:horizon] - prediction)^2)
  51                     }
  52                 }
  53                 return (error / nb_jours)
  54             }
  55
  56             if (simtype != "endo")
  57             {
  58                 h_best_exo = optimize(
  59                     errorOnLastNdays, c(0,10), kernel=kernel, simtype="exo")$minimum
  60             }
  61             if (simtype != "exo")
  62             {
  63                 h_best_endo = optimize(
  64                     errorOnLastNdays, c(0,10), kernel=kernel, simtype="endo")$minimum
  65             }
  66
  67             if (simtype == "endo")
  68             {
  69                 return (private$.predictShapeAux(data,
  70                     fdays, today, horizon, h_best_endo, kernel, "endo", TRUE))
  71             }
  72             if (simtype == "exo")
  73             {
  74                 return (private$.predictShapeAux(data,
  75                     fdays, today, horizon, h_best_exo, kernel, "exo", TRUE))
  76             }
  77             if (simtype == "mix")
  78             {
  79                 h_best_mix = c(h_best_endo,h_best_exo)
  80                 return(private$.predictShapeAux(data,
  81                     fdays, today, horizon, h_best_mix, kernel, "mix", TRUE))
  82             }
  83         }
  84     ),
  85     private = list(
  86         # Precondition: "today" is full (no NAs)
  87         .predictShapeAux = function(data, fdays, today, horizon, h, kernel, simtype, final_call)
  88         {
  89             fdays_cut = fdays[ fdays < today ]
  90             # TODO: 3 = magic number
  91             if (length(fdays_cut) < 3)
  92                 return (NA)
  93
  94             # Neighbors: days in "same season"; TODO: 60 == magic number...
  95             fdays = getSimilarDaysIndices(today, data, limit=60, same_season=TRUE, days_in=fdays_cut)
  96             if (length(fdays) <= 1)
  97                 return (NA)
  98             levelToday = data$getLevel(today)
  99             distances = sapply(fdays, function(i) abs(data$getLevel(i)-levelToday))
 100             dist_thresh = 1
 101             repeat
 102             {
 103                 same_pollution = (distances <= dist_thresh)
 104                 if (sum(same_pollution) >= 2) #will eventually happen
 105                     break
 106                 dist_thresh = dist_thresh + 1
 107             }
 108             fdays = fdays[same_pollution]
 109             if (length(fdays) == 1)
 110             {
 111                 if (final_call)
 112                 {
 113                     private$.params$weights <- 1
 114                     private$.params$indices <- fdays
 115                     private$.params$window <- 1
 116                 }
 117                 return ( data$getSerie(fdays[1])[1:horizon] ) #what else?!
 118             }
 119
 120             if (simtype != "exo")
 121             {
 122                 h_endo = ifelse(simtype=="mix", h[1], h)
 123
 124                 # Distances from last observed day to days in the past
 125                 serieToday = data$getSerie(today)
 126                 distances2 = sapply(fdays, function(i) {
 127                     delta = serieToday - data$getSerie(i)
 128                     mean(delta^2)
 129                 })
 130
 131                 sd_dist = sd(distances2)
 132                 if (sd_dist < .Machine$double.eps)
 133                 {
 134 #                   warning("All computed distances are very close: stdev too small")
 135                     sd_dist = 1 #mostly for tests... FIXME:
 136                 }
 137                 simils_endo =
 138                     if (kernel=="Gauss")
 139                         exp(-distances2/(sd_dist*h_endo^2))
 140                     else
 141                     {
 142                         # Epanechnikov
 143                         u = 1 - distances2/(sd_dist*h_endo^2)
 144                         u[abs(u)>1] = 0.
 145                         u
 146                     }
 147             }
 148
 149             if (simtype != "endo")
 150             {
 151                 h_exo = ifelse(simtype=="mix", h[2], h)
 152
 153                 M = matrix( nrow=1+length(fdays), ncol=1+length(data$getExo(today)) )
 154                 M[1,] = c( data$getLevel(today), as.double(data$getExo(today)) )
 155                 for (i in seq_along(fdays))
 156                     M[i+1,] = c( data$getLevel(fdays[i]), as.double(data$getExo(fdays[i])) )
 157
 158                 sigma = cov(M) #NOTE: robust covariance is way too slow
 159                 # TODO: 10 == magic number; more robust way == det, or always ginv()
 160                 sigma_inv =
 161                     if (length(fdays) > 10)
 162                         solve(sigma)
 163                     else
 164                         MASS::ginv(sigma)
 165
 166                 # Distances from last observed day to days in the past
 167                 distances2 = sapply(seq_along(fdays), function(i) {
 168                     delta = M[1,] - M[i+1,]
 169                     delta %*% sigma_inv %*% delta
 170                 })
 171
 172                 sd_dist = sd(distances2)
 173                 if (sd_dist < .25 * sqrt(.Machine$double.eps))
 174                 {
 175 #                   warning("All computed distances are very close: stdev too small")
 176                     sd_dist = 1 #mostly for tests... FIXME:
 177                 }
 178                 simils_exo =
 179                     if (kernel=="Gauss")
 180                         exp(-distances2/(sd_dist*h_exo^2))
 181                     else
 182                     {
 183                         # Epanechnikov
 184                         u = 1 - distances2/(sd_dist*h_exo^2)
 185                         u[abs(u)>1] = 0.
 186                         u
 187                     }
 188             }
 189
 190             similarities =
 191                 if (simtype == "exo")
 192                     simils_exo
 193                 else if (simtype == "endo")
 194                     simils_endo
 195                 else #mix
 196                     simils_endo * simils_exo
 197
 198             prediction = rep(0, horizon)
 199             for (i in seq_along(fdays))
 200                 prediction = prediction + similarities[i] * data$getCenteredSerie(fdays[i]+1)[1:horizon]
 201             prediction = prediction / sum(similarities, na.rm=TRUE)
 202
 203             if (final_call)
 204             {
 205                 private$.params$weights <- similarities
 206                 private$.params$indices <- fdays
 207                 private$.params$window <-
 208                     if (simtype=="endo")
 209                         h_endo
 210                     else if (simtype=="exo")
 211                         h_exo
 212                     else #mix
 213                         c(h_endo,h_exo)
 214             }
 215
 216             return (prediction)
 217         }
 218     )
 219 )