X-Git-Url: https://git.auder.net/?a=blobdiff_plain;f=pkg%2FR%2FF_Neighbors.R;h=d889a34ce64b82c51889a5a81322a1a7dec25b2c;hb=ee8b1b4e3c13f8dcf13a2c8da6a3bef1520c8252;hp=5b1f82659b75bb053fcb4f1bf3b10fcceded9b32;hpb=fa5b7bfc79eeffcb6e1f7b254c2d6786ba5cdbe7;p=talweg.git

diff --git a/pkg/R/F_Neighbors.R b/pkg/R/F_Neighbors.R
index 5b1f826..d889a34 100644
--- a/pkg/R/F_Neighbors.R
+++ b/pkg/R/F_Neighbors.R
@@ -31,23 +31,27 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 			}
 
 			# Indices of similar days for cross-validation; TODO: 45 = magic number
-			sdays = getSimilarDaysIndices(today, limit=45, same_season=FALSE)
+			sdays = getSimilarDaysIndices(today, data, limit=45, same_season=FALSE)
+
+			cv_days = intersect(fdays,sdays)
+			# Limit to 20 most recent matching days (TODO: 20 == magic number)
+			cv_days = sort(cv_days,decreasing=TRUE)[1:min(20,length(cv_days))]
 
 			# Function to optimize h : h |--> sum of prediction errors on last 45 "similar" days
 			errorOnLastNdays = function(h, kernel, simtype)
 			{
 				error = 0
 				nb_jours = 0
-				for (i in intersect(fdays,sdays))
+				for (i in seq_along(cv_days))
 				{
 					# mix_strategy is never used here (simtype != "mix"), therefore left blank
 					prediction = private$.predictShapeAux(data,
-						fdays, i, horizon, h, kernel, simtype, FALSE)
+						fdays, cv_days[i], horizon, h, kernel, simtype, FALSE)
 					if (!is.na(prediction[1]))
 					{
 						nb_jours = nb_jours + 1
 						error = error +
-							mean((data$getCenteredSerie(i+1)[1:horizon] - prediction)^2)
+							mean((data$getCenteredSerie(cv_days[i]+1)[1:horizon] - prediction)^2)
 					}
 				}
 				return (error / nb_jours)
@@ -96,14 +100,11 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 				h_endo = ifelse(simtype=="mix", h[1], h)
 
 				# Distances from last observed day to days in the past
-				distances2 = rep(NA, length(fdays))
-				for (i in seq_along(fdays))
-				{
-					delta = data$getCenteredSerie(today) - data$getCenteredSerie(fdays[i])
-					# Require at least half of non-NA common values to compute the distance
-					if ( !any( is.na(delta) ) )
-						distances2[i] = mean(delta^2)
-				}
+				serieToday = data$getSerie(today)
+				distances2 = sapply(fdays, function(i) {
+					delta = serieToday - data$getSerie(i)
+					mean(delta^2)
+				})
 
 				sd_dist = sd(distances2)
 				if (sd_dist < .Machine$double.eps)
@@ -133,18 +134,21 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 					M[i+1,] = c( data$getLevel(fdays[i]), as.double(data$getExo(fdays[i])) )
 
 				sigma = cov(M) #NOTE: robust covariance is way too slow
-				sigma_inv = solve(sigma) #TODO: use pseudo-inverse if needed?
+				# TODO: 10 == magic number; more robust way == det, or always ginv()
+				sigma_inv =
+					if (length(fdays) > 10)
+						solve(sigma)
+					else
+						MASS::ginv(sigma)
 
 				# Distances from last observed day to days in the past
-				distances2 = rep(NA, nrow(M)-1)
-				for (i in 2:nrow(M))
-				{
-					delta = M[1,] - M[i,]
-					distances2[i-1] = delta %*% sigma_inv %*% delta
-				}
+				distances2 = sapply(seq_along(fdays), function(i) {
+					delta = M[1,] - M[i+1,]
+					delta %*% sigma_inv %*% delta
+				})
 
 				sd_dist = sd(distances2)
-				if (sd_dist < .Machine$double.eps)
+				if (sd_dist < .25 * sqrt(.Machine$double.eps))
 				{
 #					warning("All computed distances are very close: stdev too small")
 					sd_dist = 1 #mostly for tests... FIXME:
@@ -171,7 +175,7 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster",
 
 			prediction = rep(0, horizon)
 			for (i in seq_along(fdays))
-				prediction = prediction + similarities[i] * data$getSerie(fdays[i]+1)[1:horizon]
+				prediction = prediction + similarities[i] * data$getCenteredSerie(fdays[i]+1)[1:horizon]
 			prediction = prediction / sum(similarities, na.rm=TRUE)
 
 			if (final_call)