X-Git-Url: https://git.auder.net/?a=blobdiff_plain;f=pkg%2FR%2FF_Neighbors.R;h=ea18bb6bbda5220ce975ba7c06682006a6a6d3cb;hb=d2ab47a744d8fb29c03a76a7ca2368dae53f9a57;hp=52c2b35bfb31e168fbe3b4750761c6bd04d4edfa;hpb=2057c793ad9929ed5bef8663ea28b896c84df0fc;p=talweg.git

diff --git a/pkg/R/F_Neighbors.R b/pkg/R/F_Neighbors.R
index 52c2b35..ea18bb6 100644
--- a/pkg/R/F_Neighbors.R
+++ b/pkg/R/F_Neighbors.R
@@ -1,23 +1,58 @@
 #' Neighbors Forecaster
 #'
-#' Predict tomorrow as a weighted combination of "futures of the past" days.
-#' Inherits \code{\link{Forecaster}}
+#' Predict next serie as a weighted combination of "futures of the past" days,
+#' where days in the past are chosen and weighted according to some similarity measures.
+#'
+#' The main method is \code{predictShape()}, taking arguments data, today, memory,
+#' predict_from, horizon respectively for the dataset (object output of
+#' \code{getData()}), the current index, the data depth (in days), the first predicted
+#' hour and the last predicted hour.
+#' In addition, optional arguments can be passed:
+#' \itemize{
+#'   \item local : TRUE (default) to constrain neighbors to be "same days within same
+#'     season"
+#'   \item simtype : 'endo' for a similarity based on the series only,<cr>
+#'             'exo' for a similarity based on exogenous variables only,<cr>
+#'             'mix' for the product of 'endo' and 'exo',<cr>
+#'             'none' (default) to apply a simple average: no computed weights
+#'   \item window : A window for similarities computations; override cross-validation
+#'     window estimation.
+#' }
+#' The method is summarized as follows:
+#' \enumerate{
+#'   \item Determine N (=20) recent days without missing values, and followed by a
+#'     tomorrow also without missing values.
+#'   \item Optimize the window parameters (if relevant) on the N chosen days.
+#'   \item Considering the optimized window, compute the neighbors (with locality
+#'     constraint or not), compute their similarities -- using a gaussian kernel if
+#'     simtype != "none" -- and average accordingly the "tomorrows of neigbors" to
+#'     obtain the final prediction.
+#' }
+#'
+#' @usage # NeighborsForecaster$new(pjump)
+#'
+#' @docType class
+#' @format R6 class, inherits Forecaster
+#' @aliases F_Neighbors
 #'
 NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 	inherit = Forecaster,
 
 	public = list(
-		predictShape = function(data, today, memory, horizon, ...)
+		predictShape = function(data, today, memory, predict_from, horizon, ...)
 		{
 			# (re)initialize computed parameters
 			private$.params <- list("weights"=NA, "indices"=NA, "window"=NA)
 
 			# Do not forecast on days with NAs (TODO: softer condition...)
-			if (any(is.na(data$getCenteredSerie(today))))
+			if (any(is.na(data$getSerie(today-1)))
+				|| any(is.na(data$getSerie(today)[1:(predict_from-1)])))
+			{
 				return (NA)
+			}
 
 			# Determine indices of no-NAs days followed by no-NAs tomorrows
-			fdays = getNoNA2(data, max(today-memory,1), today-1)
+			fdays = .getNoNA2(data, max(today-memory,1), today-2)
 
 			# Get optional args
 			local = ifelse(hasArg("local"), list(...)$local, TRUE) #same level + season?
@@ -25,7 +60,7 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 			if (hasArg("window"))
 			{
 				return ( private$.predictShapeAux(data,
-					fdays, today, horizon, local, list(...)$window, simtype, TRUE) )
+					fdays, today, predict_from, horizon, local, list(...)$window, simtype, TRUE) )
 			}
 
 			# Indices of similar days for cross-validation; TODO: 20 = magic number
@@ -40,13 +75,13 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 				for (i in seq_along(cv_days))
 				{
 					# mix_strategy is never used here (simtype != "mix"), therefore left blank
-					prediction = private$.predictShapeAux(data, fdays, cv_days[i], horizon, local,
-						window, simtype, FALSE)
+					prediction = private$.predictShapeAux(data, fdays, cv_days[i], predict_from,
+						horizon, local, window, simtype, FALSE)
 					if (!is.na(prediction[1]))
 					{
 						nb_jours = nb_jours + 1
 						error = error +
-							mean((data$getSerie(cv_days[i]+1)[1:horizon] - prediction)^2)
+							mean((data$getSerie(cv_days[i]+1)[predict_from:horizon] - prediction)^2)
 					}
 				}
 				return (error / nb_jours)
@@ -74,14 +109,14 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 				else #none: value doesn't matter
 					1
 
-			return(private$.predictShapeAux(data, fdays, today, horizon, local,
-				best_window, simtype, TRUE))
+			return( private$.predictShapeAux(data, fdays, today, predict_from, horizon, local,
+				best_window, simtype, TRUE) )
 		}
 	),
 	private = list(
 		# Precondition: "today" is full (no NAs)
-		.predictShapeAux = function(data, fdays, today, horizon, local, window, simtype,
-			final_call)
+		.predictShapeAux = function(data, fdays, today, predict_from, horizon, local, window,
+			simtype, final_call)
 		{
 			fdays_cut = fdays[ fdays < today ]
 			if (length(fdays_cut) <= 1)
@@ -89,33 +124,14 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 
 			if (local)
 			{
-				# Neighbors: days in "same season"; TODO: 60 == magic number...
+				# TODO: 60 == magic number
 				fdays = getSimilarDaysIndices(today, data, limit=60, same_season=TRUE,
 					days_in=fdays_cut)
 				if (length(fdays) <= 1)
 					return (NA)
-				levelToday = data$getLevel(today)
-				distances = sapply(fdays, function(i) abs(data$getLevel(i)-levelToday))
-				#TODO: 2, 10, 3, 12 magic numbers here...
-				dist_thresh = 2
-				min_neighbs = min(10,length(fdays))
-				repeat
-				{
-					same_pollution = (distances <= dist_thresh)
-					nb_neighbs = sum(same_pollution)
-					if (nb_neighbs >= min_neighbs) #will eventually happen
-						break
-					dist_thresh = dist_thresh + 3
-				}
-				fdays = fdays[same_pollution]
-				max_neighbs = 12
-				if (nb_neighbs > max_neighbs)
-				{
-					# Keep only max_neighbs closest neighbors
-					fdays = fdays[
-						sort(distances[same_pollution],index.return=TRUE)$ix[1:max_neighbs] ]
-				}
-				if (length(fdays) == 1) #the other extreme...
+				# TODO: 10, 12 == magic numbers
+				fdays = .getConstrainedNeighbs(today,data,fdays,min_neighbs=10,max_neighbs=12)
+				if (length(fdays) == 1)
 				{
 					if (final_call)
 					{
@@ -123,7 +139,7 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 						private$.params$indices <- fdays
 						private$.params$window <- 1
 					}
-					return ( data$getSerie(fdays[1])[1:horizon] )
+					return ( data$getSerie(fdays[1]+1)[predict_from:horizon] )
 				}
 			}
 			else
@@ -135,19 +151,13 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 				window_endo = ifelse(simtype=="mix", window[1], window)
 
 				# Distances from last observed day to days in the past
-				serieToday = data$getSerie(today)
+				lastSerie = c( data$getSerie(today-1), data$getSerie(today)[1:(predict_from-1)] )
 				distances2 = sapply(fdays, function(i) {
-					delta = serieToday - data$getSerie(i)
-					mean(delta^2)
+					delta = lastSerie - c(data$getSerie(i),data$getSerie(i+1)[1:(predict_from-1)])
+					sqrt(mean(delta^2))
 				})
 
-				sd_dist = sd(distances2)
-				if (sd_dist < .25 * sqrt(.Machine$double.eps))
-				{
-#					warning("All computed distances are very close: stdev too small")
-					sd_dist = 1 #mostly for tests... FIXME:
-				}
-				simils_endo = exp(-distances2/(sd_dist*window_endo^2))
+				simils_endo <- .computeSimils(distances2, window_endo)
 			}
 
 			if (simtype == "exo" || simtype == "mix")
@@ -155,12 +165,12 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 				# Compute exogen similarities using given window
 				window_exo = ifelse(simtype=="mix", window[2], window)
 
-				M = matrix( nrow=1+length(fdays), ncol=1+length(data$getExo(today)) )
-				M[1,] = c( data$getLevel(today), as.double(data$getExo(today)) )
+				M = matrix( ncol=1+length(fdays), nrow=1+length(data$getExo(1)) )
+				M[,1] = c( data$getLevelHat(today), as.double(data$getExoHat(today)) )
 				for (i in seq_along(fdays))
-					M[i+1,] = c( data$getLevel(fdays[i]), as.double(data$getExo(fdays[i])) )
+					M[,i+1] = c( data$getLevel(fdays[i]), as.double(data$getExo(fdays[i])) )
 
-				sigma = cov(M) #NOTE: robust covariance is way too slow
+				sigma = cov(t(M)) #NOTE: robust covariance is way too slow
 				# TODO: 10 == magic number; more robust way == det, or always ginv()
 				sigma_inv =
 					if (length(fdays) > 10)
@@ -170,17 +180,11 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 
 				# Distances from last observed day to days in the past
 				distances2 = sapply(seq_along(fdays), function(i) {
-					delta = M[1,] - M[i+1,]
+					delta = M[,1] - M[,i+1]
 					delta %*% sigma_inv %*% delta
 				})
 
-				sd_dist = sd(distances2)
-				if (sd_dist < .25 * sqrt(.Machine$double.eps))
-				{
-#					warning("All computed distances are very close: stdev too small")
-					sd_dist = 1 #mostly for tests... FIXME:
-				}
-				simils_exo = exp(-distances2/(sd_dist*window_exo^2))
+				simils_exo <- .computeSimils(distances2, window_exo)
 			}
 
 			similarities =
@@ -194,9 +198,12 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 					rep(1, length(fdays))
 			similarities = similarities / sum(similarities)
 
-			prediction = rep(0, horizon)
+			prediction = rep(0, horizon-predict_from+1)
 			for (i in seq_along(fdays))
-				prediction = prediction + similarities[i] * data$getSerie(fdays[i]+1)[1:horizon]
+			{
+				prediction = prediction +
+					similarities[i] * data$getSerie(fdays[i]+1)[predict_from:horizon]
+			}
 
 			if (final_call)
 			{
@@ -217,3 +224,59 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 		}
 	)
 )
+
+# getConstrainedNeighbs
+#
+# Get indices of neighbors of similar pollution level (among same season + day type).
+#
+# @param today Index of current day
+# @param data Object of class Data
+# @param fdays Current set of "first days" (no-NA pairs)
+# @param min_neighbs Minimum number of points in a neighborhood
+# @param max_neighbs Maximum number of points in a neighborhood
+#
+.getConstrainedNeighbs = function(today, data, fdays, min_neighbs=10, max_neighbs=12)
+{
+	levelToday = data$getLevelHat(today)
+	levelYersteday = data$getLevel(today-1)
+	distances = sapply(fdays, function(i) {
+		sqrt((data$getLevel(i)-levelYersteday)^2 + (data$getLevel(i+1)-levelToday)^2)
+	})
+	#TODO: 1, +1, +3 : magic numbers
+	dist_thresh = 1
+	min_neighbs = min(min_neighbs,length(fdays))
+	repeat
+	{
+		same_pollution = (distances <= dist_thresh)
+		nb_neighbs = sum(same_pollution)
+		if (nb_neighbs >= min_neighbs) #will eventually happen
+			break
+		dist_thresh = dist_thresh + ifelse(dist_thresh>1,3,1)
+	}
+	fdays = fdays[same_pollution]
+	max_neighbs = 12
+	if (nb_neighbs > max_neighbs)
+	{
+		# Keep only max_neighbs closest neighbors
+		fdays = fdays[ order(distances[same_pollution])[1:max_neighbs] ]
+	}
+	fdays
+}
+
+# compute similarities
+#
+# Apply the gaussian kernel on computed squared distances.
+#
+# @param distances2 Squared distances
+# @param window Window parameter for the kernel
+#
+.computeSimils <- function(distances2, window)
+{
+	sd_dist = sd(distances2)
+	if (sd_dist < .25 * sqrt(.Machine$double.eps))
+	{
+#		warning("All computed distances are very close: stdev too small")
+		sd_dist = 1 #mostly for tests... FIXME:
+	}
+	exp(-distances2/(sd_dist*window^2))
+}