From: Benjamin Auder Date: Wed, 29 Mar 2017 01:06:08 +0000 (+0200) Subject: No longer direct predict for Neighbors2: recollement comme Neighbors (better) X-Git-Url: https://git.auder.net/doc/html/css/current/scripts/pieces/DESCRIPTION?a=commitdiff_plain;h=6774e53de7b8bdac191d6203a380ad46c3b4d9ba;p=talweg.git No longer direct predict for Neighbors2: recollement comme Neighbors (better) --- diff --git a/pkg/R/F_Neighbors.R b/pkg/R/F_Neighbors.R index d889a34..c8a3355 100644 --- a/pkg/R/F_Neighbors.R +++ b/pkg/R/F_Neighbors.R @@ -30,12 +30,8 @@ NeighborsForecaster = R6::R6Class("NeighborsForecaster", fdays, today, horizon, list(...)$h_window, kernel, simtype, TRUE) ) } - # Indices of similar days for cross-validation; TODO: 45 = magic number - sdays = getSimilarDaysIndices(today, data, limit=45, same_season=FALSE) - - cv_days = intersect(fdays,sdays) - # Limit to 20 most recent matching days (TODO: 20 == magic number) - cv_days = sort(cv_days,decreasing=TRUE)[1:min(20,length(cv_days))] + # Indices of similar days for cross-validation; TODO: 20 = magic number + cv_days = getSimilarDaysIndices(today, data, limit=20, same_season=FALSE, days_in=fdays) # Function to optimize h : h |--> sum of prediction errors on last 45 "similar" days errorOnLastNdays = function(h, kernel, simtype) diff --git a/pkg/R/F_Neighbors2.R b/pkg/R/F_Neighbors2.R index 787dd2b..69e69dc 100644 --- a/pkg/R/F_Neighbors2.R +++ b/pkg/R/F_Neighbors2.R @@ -9,11 +9,6 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", inherit = Forecaster, public = list( - predictSerie = function(data, today, memory, horizon, ...) - { - # This method predict shape + level at the same time, all in next call - self$predictShape(data, today, memory, horizon, ...) - }, predictShape = function(data, today, memory, horizon, ...) { # (re)initialize computed parameters @@ -35,12 +30,8 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", fdays, today, horizon, list(...)$h_window, kernel, simtype, TRUE) ) } - # Indices of similar days for cross-validation; TODO: 45 = magic number - sdays = getSimilarDaysIndices(today, data, limit=45, same_season=FALSE) - - cv_days = intersect(fdays,sdays) - # Limit to 20 most recent matching days (TODO: 20 == magic number) - cv_days = sort(cv_days,decreasing=TRUE)[1:min(20,length(cv_days))] + # Indices of similar days for cross-validation; TODO: 20 = magic number + cv_days = getSimilarDaysIndices(today, data, limit=20, same_season=FALSE, days_in=fdays) # Function to optimize h : h |--> sum of prediction errors on last 45 "similar" days errorOnLastNdays = function(h, kernel, simtype) @@ -56,7 +47,7 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", { nb_jours = nb_jours + 1 error = error + - mean((data$getSerie(cv_days[i]+1)[1:horizon] - prediction)^2) + mean((data$getCenteredSerie(cv_days[i]+1)[1:horizon] - prediction)^2) } } return (error / nb_jours) @@ -95,36 +86,35 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", # Precondition: "today" is full (no NAs) .predictShapeAux = function(data, fdays, today, horizon, h, kernel, simtype, final_call) { - fdays = fdays[ fdays < today ] + fdays_cut = fdays[ fdays < today ] # TODO: 3 = magic number - if (length(fdays) < 3) + if (length(fdays_cut) < 3) return (NA) - # Neighbors: days in "same season" - sdays = getSimilarDaysIndices(today, data, limit=45, same_season=TRUE) - indices = intersect(fdays,sdays) - if (length(indices) <= 1) + # Neighbors: days in "same season"; TODO: 60 == magic number... + fdays = getSimilarDaysIndices(today, data, limit=60, same_season=TRUE, days_in=fdays_cut) + if (length(fdays) <= 1) return (NA) levelToday = data$getLevel(today) - distances = sapply(indices, function(i) abs(data$getLevel(i)-levelToday)) - # 2 and 5 below == magic numbers (determined by Bruno & Michel) - same_pollution = (distances <= 2) - if (sum(same_pollution) == 0) + distances = sapply(fdays, function(i) abs(data$getLevel(i)-levelToday)) + dist_thresh = 1 + repeat { - same_pollution = (distances <= 5) - if (sum(same_pollution) == 0) - return (NA) + same_pollution = (distances <= dist_thresh) + if (sum(same_pollution) >= 2) #will eventually happen + break + dist_thresh = dist_thresh + 1 } - indices = indices[same_pollution] - if (length(indices) == 1) + fdays = fdays[same_pollution] + if (length(fdays) == 1) { if (final_call) { private$.params$weights <- 1 - private$.params$indices <- indices + private$.params$indices <- fdays private$.params$window <- 1 } - return ( data$getSerie(indices[1])[1:horizon] ) #what else?! + return ( data$getSerie(fdays[1])[1:horizon] ) #what else?! } if (simtype != "exo") @@ -133,7 +123,7 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", # Distances from last observed day to days in the past serieToday = data$getSerie(today) - distances2 = sapply(indices, function(i) { + distances2 = sapply(fdays, function(i) { delta = serieToday - data$getSerie(i) mean(delta^2) }) @@ -160,21 +150,21 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", { h_exo = ifelse(simtype=="mix", h[2], h) - M = matrix( nrow=1+length(indices), ncol=1+length(data$getExo(today)) ) + M = matrix( nrow=1+length(fdays), ncol=1+length(data$getExo(today)) ) M[1,] = c( data$getLevel(today), as.double(data$getExo(today)) ) - for (i in seq_along(indices)) - M[i+1,] = c( data$getLevel(indices[i]), as.double(data$getExo(indices[i])) ) + for (i in seq_along(fdays)) + M[i+1,] = c( data$getLevel(fdays[i]), as.double(data$getExo(fdays[i])) ) sigma = cov(M) #NOTE: robust covariance is way too slow # TODO: 10 == magic number; more robust way == det, or always ginv() sigma_inv = - if (length(indices) > 10) + if (length(fdays) > 10) solve(sigma) else MASS::ginv(sigma) # Distances from last observed day to days in the past - distances2 = sapply(seq_along(indices), function(i) { + distances2 = sapply(seq_along(fdays), function(i) { delta = M[1,] - M[i+1,] delta %*% sigma_inv %*% delta }) @@ -206,14 +196,14 @@ Neighbors2Forecaster = R6::R6Class("Neighbors2Forecaster", simils_endo * simils_exo prediction = rep(0, horizon) - for (i in seq_along(indices)) - prediction = prediction + similarities[i] * data$getSerie(indices[i]+1)[1:horizon] + for (i in seq_along(fdays)) + prediction = prediction + similarities[i] * data$getCenteredSerie(fdays[i]+1)[1:horizon] prediction = prediction / sum(similarities, na.rm=TRUE) if (final_call) { private$.params$weights <- similarities - private$.params$indices <- indices + private$.params$indices <- fdays private$.params$window <- if (simtype=="endo") h_endo diff --git a/pkg/R/utils.R b/pkg/R/utils.R index b3e66e1..3649f58 100644 --- a/pkg/R/utils.R +++ b/pkg/R/utils.R @@ -59,9 +59,10 @@ integerIndexToDate = function(index, data) #' @param data Reference dataset, object output of \code{getData} #' @param limit Maximum number of indices to return #' @param same_season Should the indices correspond to day in same season? +#' @param days_in Optional set to intersect with results (NULL to discard) #' #' @export -getSimilarDaysIndices = function(index, data, limit, same_season) +getSimilarDaysIndices = function(index, data, limit, same_season, days_in=NULL) { index = dateIndexToInteger(index, data) @@ -74,7 +75,7 @@ getSimilarDaysIndices = function(index, data, limit, same_season) while (i >= 1 && length(days) < limit) { dt = as.POSIXlt(data$getTime(i)[1]) - if (.isSameDay(dt$wday, day_ref)) + if ((is.null(days_in) || i %in% days_in) && .isSameDay(dt$wday, day_ref)) { if (!same_season || .isSameSeason(dt$mon+1, month_ref)) days = c(days, i) diff --git a/reports/report.ipynb b/reports/report.ipynb index f0a06d0..8af4acd 100644 --- a/reports/report.ipynb +++ b/reports/report.ipynb @@ -12,15 +12,14 @@ "

Introduction

\n", "\n", "J'ai fait quelques essais dans différentes configurations pour la méthode \"Neighbors\"\n", - "(la seule dont on a parlé).
Il semble que le mieux soit\n", + "(la seule dont on a parlé) et sa variante récente appelée pour l'instant \"Neighbors2\",\n", + "avec simtype=\"mix\" : deux types de similarités prises en compte, puis multiplication des poids.\n", + "Pour Neighbors on prédit le saut (par la moyenne pondérée des sauts passés), et pour Neighbors2\n", + "on n'effectue aucun raccordement (prévision directe).\n", "\n", - " * simtype=\"exo\" ou \"mix\" : similarités exogènes avec/sans endogènes (fenêtre optimisée par VC)\n", - " * same_season=FALSE : les indices pour la validation croisée ne tiennent pas compte des saisons\n", - " * mix_strategy=\"mult\" : on multiplie les poids (au lieu d'en éteindre)\n", - "\n", - "J'ai systématiquement comparé à une approche naïve : la moyennes des lendemains des jours\n", - "\"similaires\" dans tout le passé ; à chaque fois sans prédiction du saut (sauf pour Neighbors :\n", - "prédiction basée sur les poids calculés).\n", + "J'ai systématiquement comparé à une approche naïve : la moyenne des lendemains des jours\n", + "\"similaires\" dans tout le passé, ainsi qu'à la persistence -- reproduisant le jour courant ou\n", + "allant chercher le futur similaire une semaine avant.\n", "\n", "Ensuite j'affiche les erreurs, quelques courbes prévues/mesurées, quelques filaments puis les\n", "histogrammes de quelques poids. Concernant les graphes de filaments, la moitié gauche du graphe\n", @@ -41,15 +40,18 @@ "source": [ "library(talweg)\n", "\n", + "P = 7 #instant de prévision\n", + "H = 17 #horizon (en heures)\n", + "\n", "ts_data = read.csv(system.file(\"extdata\",\"pm10_mesures_H_loc_report.csv\",package=\"talweg\"))\n", "exo_data = read.csv(system.file(\"extdata\",\"meteo_extra_noNAs.csv\",package=\"talweg\"))\n", - "# Predict from P+1 to P+H included\n", - "H = 17\n", - "data = getData(ts_data, exo_data, input_tz = \"GMT\", working_tz=\"GMT\", predict_at=7)\n", + "# NOTE: 'GMT' because DST gaps are filled and multiple values merged in above dataset.\n", + "# Prediction from P+1 to P+H included.\n", + "data = getData(ts_data, exo_data, input_tz = \"GMT\", working_tz=\"GMT\", predict_at=P)\n", "\n", "indices_ch = seq(as.Date(\"2015-01-18\"),as.Date(\"2015-01-24\"),\"days\")\n", "indices_ep = seq(as.Date(\"2015-03-15\"),as.Date(\"2015-03-21\"),\"days\")\n", - "indices_np = seq(as.Date(\"2015-04-26\"),as.Date(\"2015-05-02\"),\"days\")" + "indices_np = seq(as.Date(\"2015-04-26\"),as.Date(\"2015-05-02\"),\"days\")\n" ] }, { @@ -59,6 +61,8 @@ "editable": true }, "source": [ + "\n", + "\n", "

Pollution par chauffage

" ] }, @@ -73,58 +77,21 @@ "outputs": [], "source": [ "reload(\"../pkg\")\n", - "#p1 = computeForecast(data, indices_ch, \"Neighbors\", \"Zero\", horizon=H, simtype=\"exo\")\n", - "#p2 = computeForecast(data, indices_ch, \"Neighbors\", \"Zero\", horizon=H, simtype=\"endo\")\n", - "p3 = computeForecast(data, indices_ch, \"Neighbors\", \"Zero\", horizon=H, simtype=\"mix\")\n", - "p4 = computeForecast(data, indices_ch, \"Neighbors\", \"Neighbors\", horizon=H, simtype=\"mix\")\n", - "#p4 = computeForecast(data, indices_ch, \"Neighbors2\", \"Zero\", horizon=H, simtype=\"exo\")\n", - "#p5 = computeForecast(data, indices_ch, \"Neighbors2\", \"Zero\", horizon=H, simtype=\"endo\")\n", - "#p6 = computeForecast(data, indices_ch, \"Neighbors2\", \"Zero\", horizon=H, simtype=\"mix\")" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "deletable": true, - "editable": true - }, - "outputs": [], - "source": [ - "getSimilarDaysIndices(1000,10,TRUE,data)" + "p_nn = computeForecast(data, indices_ch, \"Neighbors\", \"Neighbors\", horizon=H)\n", + "p_nn2 = computeForecast(data, indices_ch, \"Neighbors2\", \"Neighbors\", horizon=H)\n", + "p_az = computeForecast(data, indices_ch, \"Average\", \"Zero\", horizon=H)\n", + "p_pz = computeForecast(data, indices_ch, \"Persistence\", \"Zero\", horizon=H, same_day=TRUE)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { - "collapsed": false, - "deletable": true, - "editable": true - }, - "outputs": [], - "source": [ - "as.POSIXlt(data$getTime(1000)[1])" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false, - "deletable": true, - "editable": true + "collapsed": false }, "outputs": [], "source": [ - "#e1 = computeError(data, p1, H)\n", - "#e2 = computeError(data, p2, H)\n", - "e3 = computeError(data, p3, H)\n", - "e4 = computeError(data, p4, H)\n", - "#e5 = computeError(data, p5, H)\n", - "#e6 = computeError(data, p6, H)\n", - "plotError(list(e3,e4), cols=c(1,2))" + "p_nn2$getParams(5)$weights" ] }, { @@ -137,30 +104,17 @@ }, "outputs": [], "source": [ - "\tfirst_day = 1\n", - "params=p3$getParams(3)\n", - "\tfilter = (params$indices >= first_day)\n", - "\tindices = params$indices[filter]\n", - "\tweights = params$weights[filter]\n", - "\n", - "\n", - "\tgaps = sapply(indices, function(i) {\n", - "\t\tdata$getSerie(i+1)[1] - tail(data$getSerie(i), 1)\n", - "\t})\n", - "\tscal_product = weights * gaps\n", - "\tnorm_fact = sum( weights[!is.na(scal_product)] )\n", - "\tsum(scal_product, na.rm=TRUE) / norm_fact\n" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": false - }, - "outputs": [], - "source": [ - "hist(weights)" + "e_nn = computeError(data, p_nn, H)\n", + "e_nn2 = computeError(data, p_nn2, H)\n", + "e_az = computeError(data, p_az, H)\n", + "e_pz = computeError(data, p_pz, H)\n", + "options(repr.plot.width=9, repr.plot.height=7)\n", + "plotError(list(e_nn, e_pz, e_az, e_nn2), cols=c(1,2,colors()[258], 4))\n", + "\n", + "# Noir: Neighbors, bleu: Neighbors2, vert: moyenne, rouge: persistence\n", + "\n", + "i_np = which.min(e_nn$abs$indices)\n", + "i_p = which.max(e_nn$abs$indices)" ] }, { @@ -176,11 +130,14 @@ "options(repr.plot.width=9, repr.plot.height=4)\n", "par(mfrow=c(1,2))\n", "\n", - "plotPredReal(data, p3, 3); title(paste(\"PredReal nn exo day\",3))\n", - "plotPredReal(data, p3, 5); title(paste(\"PredReal nn exo day\",5))\n", + "plotPredReal(data, p_nn, i_np); title(paste(\"PredReal nn day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn day\",i_p))\n", + "\n", + "plotPredReal(data, p_nn2, i_np); title(paste(\"PredReal nn2 day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn2 day\",i_p))\n", "\n", - "plotPredReal(data, p4, 3); title(paste(\"PredReal nn mix day\",3))\n", - "plotPredReal(data, p4, 5); title(paste(\"PredReal nn mix day\",5))\n", + "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", + "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", "\n", "# Bleu: prévue, noir: réalisée" ] @@ -196,11 +153,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", - "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", + "f_np = computeFilaments(data, p_nn, i_np, plot=TRUE); title(paste(\"Filaments nn day\",i_np))\n", + "f_p = computeFilaments(data, p_nn, i_p, plot=TRUE); title(paste(\"Filaments nn day\",i_p))\n", "\n", - "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", - "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" + "f_np2 = computeFilaments(data, p_nn2, i_np, plot=TRUE); title(paste(\"Filaments nn2 day\",i_np))\n", + "f_p2 = computeFilaments(data, p_nn2, i_p, plot=TRUE); title(paste(\"Filaments nn2 day\",i_p))" ] }, { @@ -214,11 +171,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", - "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", + "plotFilamentsBox(data, f_np); title(paste(\"FilBox nn day\",i_np))\n", + "plotFilamentsBox(data, f_p); title(paste(\"FilBox nn day\",i_p))\n", "\n", - "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", - "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" + "plotFilamentsBox(data, f_np2); title(paste(\"FilBox nn2 day\",i_np))\n", + "plotFilamentsBox(data, f_p2); title(paste(\"FilBox nn2 day\",i_p))" ] }, { @@ -232,11 +189,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", - "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", + "plotRelVar(data, f_np); title(paste(\"StdDev nn day\",i_np))\n", + "plotRelVar(data, f_p); title(paste(\"StdDev nn day\",i_p))\n", "\n", - "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", - "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", + "plotRelVar(data, f_np2); title(paste(\"StdDev nn2 day\",i_np))\n", + "plotRelVar(data, f_p2); title(paste(\"StdDev nn2 day\",i_p))\n", "\n", "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" ] @@ -252,11 +209,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", - "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", + "plotSimils(p_nn, i_np); title(paste(\"Weights nn day\",i_np))\n", + "plotSimils(p_nn, i_p); title(paste(\"Weights nn day\",i_p))\n", "\n", - "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", - "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", + "plotSimils(p_nn2, i_np); title(paste(\"Weights nn2 day\",i_np))\n", + "plotSimils(p_nn2, i_p); title(paste(\"Weights nn2 day\",i_p))\n", "\n", "# - pollué à gauche, + pollué à droite" ] @@ -271,12 +228,12 @@ }, "outputs": [], "source": [ - "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", - "p_nn_exo$getParams(i_np)$window\n", - "p_nn_exo$getParams(i_p)$window\n", + "# Fenêtres sélectionnées dans ]0,7] / nn à gauche, nn2 à droite\n", + "p_nn$getParams(i_np)$window\n", + "p_nn$getParams(i_p)$window\n", "\n", - "p_nn_mix$getParams(i_np)$window\n", - "p_nn_mix$getParams(i_p)$window" + "p_nn2$getParams(i_np)$window\n", + "p_nn2$getParams(i_p)$window" ] }, { @@ -301,14 +258,10 @@ }, "outputs": [], "source": [ - "p_nn_exo = computeForecast(data, indices_ep, \"Neighbors\", \"Neighbors\",\n", - "\thorizon=3, simtype=\"exo\")\n", - "p_nn_mix = computeForecast(data, indices_ep, \"Neighbors\", \"Neighbors\",\n", - "\thorizon=3, simtype=\"mix\")\n", - "p_az = computeForecast(data, indices_ep, \"Average\", \"Zero\",\n", - "\thorizon=3)\n", - "p_pz = computeForecast(data, indices_ep, \"Persistence\", \"Zero\",\n", - "\thorizon=3, same_day=TRUE)" + "p_nn = computeForecast(data, indices_ep, \"Neighbors\", \"Neighbors\", horizon=H)\n", + "p_nn2 = computeForecast(data, indices_ep, \"Neighbors2\", \"Zero\", horizon=H)\n", + "p_az = computeForecast(data, indices_ep, \"Average\", \"Zero\", horizon=H)\n", + "p_pz = computeForecast(data, indices_ep, \"Persistence\", \"Zero\", horizon=H, same_day=TRUE)" ] }, { @@ -321,17 +274,17 @@ }, "outputs": [], "source": [ - "e_nn_exo = computeError(data, p_nn_exo, 3)\n", - "e_nn_mix = computeError(data, p_nn_mix, 3)\n", - "e_az = computeError(data, p_az, 3)\n", - "e_pz = computeError(data, p_pz, 3)\n", + "e_nn = computeError(data, p_nn, H)\n", + "e_nn2 = computeError(data, p_nn2, H)\n", + "e_az = computeError(data, p_az, H)\n", + "e_pz = computeError(data, p_pz, H)\n", "options(repr.plot.width=9, repr.plot.height=7)\n", - "plotError(list(e_nn_mix, e_pz, e_az, e_nn_exo), cols=c(1,2,colors()[258], 4))\n", + "plotError(list(e_nn, e_pz, e_az, e_nn2), cols=c(1,2,colors()[258], 4))\n", "\n", - "# Noir: neighbors_mix, bleu: neighbors_exo, vert: moyenne, rouge: persistence\n", + "# Noir: Neighbors, bleu: Neighbors2, vert: moyenne, rouge: persistence\n", "\n", - "i_np = which.min(e_nn_exo$abs$indices)\n", - "i_p = which.max(e_nn_exo$abs$indices)" + "i_np = which.min(e_nn$abs$indices)\n", + "i_p = which.max(e_nn$abs$indices)" ] }, { @@ -347,11 +300,11 @@ "options(repr.plot.width=9, repr.plot.height=4)\n", "par(mfrow=c(1,2))\n", "\n", - "plotPredReal(data, p_nn_exo, i_np); title(paste(\"PredReal nn exo day\",i_np))\n", - "plotPredReal(data, p_nn_exo, i_p); title(paste(\"PredReal nn exo day\",i_p))\n", + "plotPredReal(data, p_nn, i_np); title(paste(\"PredReal nn day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn day\",i_p))\n", "\n", - "plotPredReal(data, p_nn_mix, i_np); title(paste(\"PredReal nn mix day\",i_np))\n", - "plotPredReal(data, p_nn_mix, i_p); title(paste(\"PredReal nn mix day\",i_p))\n", + "plotPredReal(data, p_nn2, i_np); title(paste(\"PredReal nn2 day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn2 day\",i_p))\n", "\n", "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", @@ -370,11 +323,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", - "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", + "f_np = computeFilaments(data, p_nn, i_np, plot=TRUE); title(paste(\"Filaments nn day\",i_np))\n", + "f_p = computeFilaments(data, p_nn, i_p, plot=TRUE); title(paste(\"Filaments nn day\",i_p))\n", "\n", - "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", - "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" + "f_np2 = computeFilaments(data, p_nn2, i_np, plot=TRUE); title(paste(\"Filaments nn2 day\",i_np))\n", + "f_p2 = computeFilaments(data, p_nn2, i_p, plot=TRUE); title(paste(\"Filaments nn2 day\",i_p))" ] }, { @@ -388,11 +341,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", - "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", + "plotFilamentsBox(data, f_np); title(paste(\"FilBox nn day\",i_np))\n", + "plotFilamentsBox(data, f_p); title(paste(\"FilBox nn day\",i_p))\n", "\n", - "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", - "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" + "plotFilamentsBox(data, f_np2); title(paste(\"FilBox nn2 day\",i_np))\n", + "plotFilamentsBox(data, f_p2); title(paste(\"FilBox nn2 day\",i_p))" ] }, { @@ -406,11 +359,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", - "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", + "plotRelVar(data, f_np); title(paste(\"StdDev nn day\",i_np))\n", + "plotRelVar(data, f_p); title(paste(\"StdDev nn day\",i_p))\n", "\n", - "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", - "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", + "plotRelVar(data, f_np2); title(paste(\"StdDev nn2 day\",i_np))\n", + "plotRelVar(data, f_p2); title(paste(\"StdDev nn2 day\",i_p))\n", "\n", "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" ] @@ -426,11 +379,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", - "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", + "plotSimils(p_nn, i_np); title(paste(\"Weights nn day\",i_np))\n", + "plotSimils(p_nn, i_p); title(paste(\"Weights nn day\",i_p))\n", "\n", - "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", - "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", + "plotSimils(p_nn2, i_np); title(paste(\"Weights nn2 day\",i_np))\n", + "plotSimils(p_nn2, i_p); title(paste(\"Weights nn2 day\",i_p))\n", "\n", "# - pollué à gauche, + pollué à droite" ] @@ -445,12 +398,12 @@ }, "outputs": [], "source": [ - "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", - "p_nn_exo$getParams(i_np)$window\n", - "p_nn_exo$getParams(i_p)$window\n", + "# Fenêtres sélectionnées dans ]0,7] / nn à gauche, nn2 à droite\n", + "p_nn$getParams(i_np)$window\n", + "p_nn$getParams(i_p)$window\n", "\n", - "p_nn_mix$getParams(i_np)$window\n", - "p_nn_mix$getParams(i_p)$window" + "p_nn2$getParams(i_np)$window\n", + "p_nn2$getParams(i_p)$window" ] }, { @@ -475,14 +428,10 @@ }, "outputs": [], "source": [ - "p_nn_exo = computeForecast(data, indices_np, \"Neighbors\", \"Neighbors\",\n", - "\thorizon=3, simtype=\"exo\")\n", - "p_nn_mix = computeForecast(data, indices_np, \"Neighbors\", \"Neighbors\",\n", - "\thorizon=3, simtype=\"mix\")\n", - "p_az = computeForecast(data, indices_np, \"Average\", \"Zero\",\n", - "\thorizon=3)\n", - "p_pz = computeForecast(data, indices_np, \"Persistence\", \"Zero\",\n", - "\thorizon=3, same_day=FALSE)" + "p_nn = computeForecast(data, indices_np, \"Neighbors\", \"Neighbors\", horizon=H)\n", + "p_nn2 = computeForecast(data, indices_np, \"Neighbors2\", \"Zero\", horizon=H)\n", + "p_az = computeForecast(data, indices_np, \"Average\", \"Zero\", horizon=H)\n", + "p_pz = computeForecast(data, indices_np, \"Persistence\", \"Zero\", horizon=H, same_day=FALSE)" ] }, { @@ -495,17 +444,17 @@ }, "outputs": [], "source": [ - "e_nn_exo = computeError(data, p_nn_exo, 3)\n", - "e_nn_mix = computeError(data, p_nn_mix, 3)\n", - "e_az = computeError(data, p_az, 3)\n", - "e_pz = computeError(data, p_pz, 3)\n", + "e_nn = computeError(data, p_nn, H)\n", + "e_nn2 = computeError(data, p_nn2, H)\n", + "e_az = computeError(data, p_az, H)\n", + "e_pz = computeError(data, p_pz, H)\n", "options(repr.plot.width=9, repr.plot.height=7)\n", - "plotError(list(e_nn_mix, e_pz, e_az, e_nn_exo), cols=c(1,2,colors()[258], 4))\n", + "plotError(list(e_nn, e_pz, e_az, e_nn2), cols=c(1,2,colors()[258], 4))\n", "\n", - "# Noir: neighbors_mix, bleu: neighbors_exo, vert: moyenne, rouge: persistence\n", + "# Noir: Neighbors, bleu: Neighbors2, vert: moyenne, rouge: persistence\n", "\n", - "i_np = which.min(e_nn_exo$abs$indices)\n", - "i_p = which.max(e_nn_exo$abs$indices)" + "i_np = which.min(e_nn$abs$indices)\n", + "i_p = which.max(e_nn$abs$indices)" ] }, { @@ -521,11 +470,11 @@ "options(repr.plot.width=9, repr.plot.height=4)\n", "par(mfrow=c(1,2))\n", "\n", - "plotPredReal(data, p_nn_exo, i_np); title(paste(\"PredReal nn exo day\",i_np))\n", - "plotPredReal(data, p_nn_exo, i_p); title(paste(\"PredReal nn exo day\",i_p))\n", + "plotPredReal(data, p_nn, i_np); title(paste(\"PredReal nn day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn day\",i_p))\n", "\n", - "plotPredReal(data, p_nn_mix, i_np); title(paste(\"PredReal nn mix day\",i_np))\n", - "plotPredReal(data, p_nn_mix, i_p); title(paste(\"PredReal nn mix day\",i_p))\n", + "plotPredReal(data, p_nn2, i_np); title(paste(\"PredReal nn2 day\",i_np))\n", + "plotPredReal(data, p_nn2, i_p); title(paste(\"PredReal nn2 day\",i_p))\n", "\n", "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", @@ -544,11 +493,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", - "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", + "f_np = computeFilaments(data, p_nn, i_np, plot=TRUE); title(paste(\"Filaments nn day\",i_np))\n", + "f_p = computeFilaments(data, p_nn, i_p, plot=TRUE); title(paste(\"Filaments nn day\",i_p))\n", "\n", - "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", - "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" + "f_np2 = computeFilaments(data, p_nn2, i_np, plot=TRUE); title(paste(\"Filaments nn2 day\",i_np))\n", + "f_p2 = computeFilaments(data, p_nn2, i_p, plot=TRUE); title(paste(\"Filaments nn2 day\",i_p))" ] }, { @@ -562,11 +511,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", - "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", + "plotFilamentsBox(data, f_np); title(paste(\"FilBox nn day\",i_np))\n", + "plotFilamentsBox(data, f_p); title(paste(\"FilBox nn day\",i_p))\n", "\n", - "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", - "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" + "plotFilamentsBox(data, f_np2); title(paste(\"FilBox nn2 day\",i_np))\n", + "plotFilamentsBox(data, f_p2); title(paste(\"FilBox nn2 day\",i_p))" ] }, { @@ -580,11 +529,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", - "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", + "plotRelVar(data, f_np); title(paste(\"StdDev nn day\",i_np))\n", + "plotRelVar(data, f_p); title(paste(\"StdDev nn day\",i_p))\n", "\n", - "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", - "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", + "plotRelVar(data, f_np2); title(paste(\"StdDev nn2 day\",i_np))\n", + "plotRelVar(data, f_p2); title(paste(\"StdDev nn2 day\",i_p))\n", "\n", "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" ] @@ -600,11 +549,11 @@ "outputs": [], "source": [ "par(mfrow=c(1,2))\n", - "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", - "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", + "plotSimils(p_nn, i_np); title(paste(\"Weights nn day\",i_np))\n", + "plotSimils(p_nn, i_p); title(paste(\"Weights nn day\",i_p))\n", "\n", - "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", - "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", + "plotSimils(p_nn2, i_np); title(paste(\"Weights nn2 day\",i_np))\n", + "plotSimils(p_nn2, i_p); title(paste(\"Weights nn2 day\",i_p))\n", "\n", "# - pollué à gauche, + pollué à droite" ] @@ -619,31 +568,12 @@ }, "outputs": [], "source": [ - "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", - "p_nn_exo$getParams(i_np)$window\n", - "p_nn_exo$getParams(i_p)$window\n", - "\n", - "p_nn_mix$getParams(i_np)$window\n", - "p_nn_mix$getParams(i_p)$window" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "deletable": true, - "editable": true - }, - "source": [ - "\n", - "\n", - "

Bilan

\n", - "\n", - "Problème difficile : on ne fait guère mieux qu'une naïve moyenne des lendemains des jours\n", - "similaires dans le passé, ce qui n'est pas loin de prédire une série constante égale à la\n", - "dernière valeur observée (méthode \"zéro\"). La persistence donne parfois de bons résultats\n", - "mais est trop instable (sensibilité à l'argument same_day).\n", + "# Fenêtres sélectionnées dans ]0,7] / nn à gauche, nn2 à droite\n", + "p_nn$getParams(i_np)$window\n", + "p_nn$getParams(i_p)$window\n", "\n", - "Comment améliorer la méthode ?" + "p_nn2$getParams(i_np)$window\n", + "p_nn2$getParams(i_p)$window" ] } ],