From: Benjamin Auder Date: Mon, 24 Apr 2017 22:12:17 +0000 (+0200) Subject: a few fixes X-Git-Url: https://git.auder.net/game/%7B%7B%20path%28%27fos_user_change_password%27%29%20%7D%7D?a=commitdiff_plain;h=1e8327df4e8abce5c23808be4f98037635bb2714;p=talweg.git a few fixes --- diff --git a/pkg/R/computeError.R b/pkg/R/computeError.R index 0b1771f..c3bc4f3 100644 --- a/pkg/R/computeError.R +++ b/pkg/R/computeError.R @@ -16,8 +16,8 @@ computeError = function(data, pred, predict_from, horizon=length(data$getSerie(1))) { L = pred$getSize() - mape_day = rep(0, horizon) - abs_day = rep(0, horizon) + mape_day = rep(0, horizon-predict_from+1) + abs_day = rep(0, horizon-predict_from+1) mape_indices = rep(NA, L) abs_indices = rep(NA, L) diff --git a/pkg/R/computeForecast.R b/pkg/R/computeForecast.R index e967cc7..ca8badd 100644 --- a/pkg/R/computeForecast.R +++ b/pkg/R/computeForecast.R @@ -71,27 +71,20 @@ computeForecast = function(data, indices, forecaster, pjump, predict_from, forecaster = forecaster_class_name$new( #.pjump = getFromNamespace(paste("get",pjump,"JumpPredict",sep=""), "talweg")) - if (ncores > 1 && requireNamespace("parallel",quietly=TRUE)) + computeOneForecast <- function(i) { - p <- parallel::mclapply(seq_along(integer_indices), function(i) { - list( - "forecast" = forecaster$predictSerie( - data, integer_indices[i], memory, predict_from, horizon, ...), - "params"= forecaster$getParameters(), - "index" = integer_indices[i] ) - }, mc.cores=ncores) - } - else - { - p <- lapply(seq_along(integer_indices), function(i) { - list( - "forecast" = forecaster$predictSerie( - data, integer_indices[i], memory, predict_from, horizon, ...), - "params"= forecaster$getParameters(), - "index" = integer_indices[i] ) - }) + list( + "forecast" = forecaster$predictSerie(data,i,memory,predict_from,horizon,...), + "params" = forecaster$getParameters(), + "index" = i ) } + p <- + if (ncores > 1 && requireNamespace("parallel",quietly=TRUE)) + parallel::mclapply(integer_indices, computeOneForecast, mc.cores=ncores) + else + lapply(integer_indices, computeOneForecast) + # TODO: find a way to fill pred in //... for (i in seq_along(integer_indices)) { diff --git a/reports/Experiments.gj b/reports/Experiments.gj index 0f102ad..d7ade40 100644 --- a/reports/Experiments.gj +++ b/reports/Experiments.gj @@ -2,8 +2,8 @@ # Résultats numériques Cette partie montre les résultats obtenus avec des variantes de l'algorithme décrit au -chapitre , en utilisant le package présenté à la section 3. Cet algorithme est -systématiquement comparé à deux approches naïves : +à la section 4, en utilisant le package présenté au chapitre précédent. Cet +algorithme est systématiquement comparé à deux approches naïves : * la moyenne des lendemains des jours "similaires" dans tout le passé, c'est-à-dire prédiction = moyenne de tous les mardis passés si le jour courant est un lundi. @@ -35,8 +35,8 @@ list_indices = ['indices_ch', 'indices_ep', 'indices_np'] -----r library(talweg) -P = ${P} #instant de prévision -H = ${H} #horizon (en heures) +P = ${P} #première heure de prévision +H = ${H} #dernière heure de prévision ts_data = read.csv(system.file("extdata","pm10_mesures_H_loc_report.csv", package="talweg")) @@ -44,8 +44,7 @@ exo_data = read.csv(system.file("extdata","meteo_extra_noNAs.csv", package="talweg")) # NOTE: 'GMT' because DST gaps are filled and multiple values merged in # above dataset. Prediction from P+1 to P+H included. -data = getData(ts_data, exo_data, input_tz = "GMT", working_tz="GMT", - predict_at=P) +data = getData(ts_data, exo_data) indices_ch = seq(as.Date("2015-01-18"),as.Date("2015-01-24"),"days") indices_ep = seq(as.Date("2015-03-15"),as.Date("2015-03-21"),"days") @@ -55,21 +54,22 @@ indices_np = seq(as.Date("2015-04-26"),as.Date("2015-05-02"),"days") ##

${list_titles[i]}

${"##"} ${list_titles[i]} -----r -p1 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Neighbors", horizon=H, - simtype="mix", local=FALSE) -p2 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Neighbors", horizon=H, - simtype="endo", local=TRUE) -p3 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Zero", horizon=H, - simtype="none", local=TRUE) -p4 = computeForecast(data, ${list_indices[i]}, "Average", "Zero", horizon=H) -p5 = computeForecast(data, ${list_indices[i]}, "Persistence", "Zero", horizon=H, - same_day=${'TRUE' if loop.index < 2 else 'FALSE'}) +p1 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Neighbors", predict_from=P, + horizon=H, simtype="mix", local=FALSE) +p2 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Neighbors", predict_from=P, + horizon=H, simtype="endo", local=TRUE) +p3 = computeForecast(data, ${list_indices[i]}, "Neighbors", "Zero", predict_from=P, + horizon=H, simtype="none", local=TRUE) +p4 = computeForecast(data, ${list_indices[i]}, "Average", "Zero", predict_from=P, + horizon=H) +p5 = computeForecast(data, ${list_indices[i]}, "Persistence", "Zero", predict_from=P, + horizon=H, same_day=${'TRUE' if loop.index < 2 else 'FALSE'}) -----r -e1 = computeError(data, p1, H) -e2 = computeError(data, p2, H) -e3 = computeError(data, p3, H) -e4 = computeError(data, p4, H) -e5 = computeError(data, p5, H) +e1 = computeError(data, p1, P, H) +e2 = computeError(data, p2, P, H) +e3 = computeError(data, p3, P, H) +e4 = computeError(data, p4, P, H) +e5 = computeError(data, p5, P, H) options(repr.plot.width=9, repr.plot.height=7) plotError(list(e1, e5, e4, e2, e3), cols=c(1,2,colors()[258],4,6)) @@ -134,14 +134,14 @@ journée sur la courbes "difficile à prévoir". % endif -----r par(mfrow=c(1,2)) -f_np1 = computeFilaments(data, p1, i_np, plot=TRUE) +f_np1 = computeFilaments(data, p1, i_np, predict_from=P, plot=TRUE) title(paste("Filaments p1 day",i_np)) -f_p1 = computeFilaments(data, p1, i_p, plot=TRUE) +f_p1 = computeFilaments(data, p1, i_p, predict_from=P, plot=TRUE) title(paste("Filaments p1 day",i_p)) -f_np2 = computeFilaments(data, p2, i_np, plot=TRUE) +f_np2 = computeFilaments(data, p2, i_np, predict_from=P, plot=TRUE) title(paste("Filaments p2 day",i_np)) -f_p2 = computeFilaments(data, p2, i_p, plot=TRUE) +f_p2 = computeFilaments(data, p2, i_p, predict_from=P, plot=TRUE) title(paste("Filaments p2 day",i_p)) ----- % if i == 0: @@ -161,8 +161,8 @@ de variabilité relative. % endif -----r par(mfrow=c(1,2)) -plotFilamentsBox(data, f_np1); title(paste("FilBox p1 day",i_np)) -plotFilamentsBox(data, f_p1); title(paste("FilBox p1 day",i_p)) +plotFilamentsBox(data, f_np1, predict_from=P); title(paste("FilBox p1 day",i_np)) +plotFilamentsBox(data, f_p1, predict_from=P); title(paste("FilBox p1 day",i_p)) # En pointillés la courbe du jour courant + lendemain (à prédire) ----- @@ -185,11 +185,11 @@ lendemains de voisins atypiques, courbe à prévoir elle-même légèrement % endif -----r par(mfrow=c(1,2)) -plotRelVar(data, f_np1); title(paste("StdDev p1 day",i_np)) -plotRelVar(data, f_p1); title(paste("StdDev p1 day",i_p)) +plotRelVar(data, f_np1, predict_from=P); title(paste("StdDev p1 day",i_np)) +plotRelVar(data, f_p1, predict_from=P); title(paste("StdDev p1 day",i_p)) -plotRelVar(data, f_np2); title(paste("StdDev p2 day",i_np)) -plotRelVar(data, f_p2); title(paste("StdDev p2 day",i_p)) +plotRelVar(data, f_np2, predict_from=P); title(paste("StdDev p2 day",i_np)) +plotRelVar(data, f_p2, preidct_from=P); title(paste("StdDev p2 day",i_p)) # Variabilité globale en rouge ; sur les voisins (+ lendemains) en noir ----- diff --git a/reports/PackageR.gj b/reports/PackageR.gj index 567bfd6..29c4822 100644 --- a/reports/PackageR.gj +++ b/reports/PackageR.gj @@ -21,17 +21,14 @@ partie suivante. library(talweg) # Acquisition des données (depuis les fichiers CSV) -ts_data <- read.csv(system.file("extdata","pm10_mesures_H_loc.csv", - package="talweg")) -exo_data <- read.csv(system.file("extdata","meteo_extra_noNAs.csv", - package="talweg")) -data <- getData(ts_data, exo_data, - date_format="%d/%m/%Y %H:%M", limit=120) +ts_data <- read.csv(system.file("extdata","pm10_mesures_H_loc.csv", package="talweg")) +exo_data <- read.csv(system.file("extdata","meteo_extra_noNAs.csv", package="talweg")) +data <- getData(ts_data, exo_data, date_format="%d/%m/%Y %H:%M", limit=120) # Plus de détails à la section 1 ci-après. # Prédiction de 10 courbes (jours 102 à 111) -pred <- computeForecast(data, 101:110, "Persistence", "Zero", - predict_from=8, memory=50, horizon=24, ncores=1) +pred <- computeForecast(data, 101:110, "Persistence", "Zero", predict_from=8, memory=50, + horizon=24, ncores=1) # Plus de détails à la section 2 ci-après. # Calcul des erreurs (sur un horizon arbitraire <= horizon de prédiction)