R version 2.9.0 (2009-04-17) Copyright (C) 2009 The R Foundation for Statistical Computing ISBN 3-900051-07-0 R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type 'license()' or 'licence()' for distribution details. R is a collaborative project with many contributors. Type 'contributors()' for more information and 'citation()' on how to cite R or R packages in publications. Type 'demo()' for some demos, 'help()' for on-line help, or 'help.start()' for an HTML browser interface to help. Type 'q()' to quit R. > x <- c(42.13333333,43.68333333,9.016666667,33.7,19.09866667,10.06733333,18.81666667,40.85,51.575,16.625,42.6,25.66666667,49.325,39.40833333,37.55333333,57.05,29.575) > par10 = '0.1' > par9 = '3' > par8 = 'dumresult' > par7 = 'dum' > par6 = '12' > par5 = 'ZZZ' > par4 = 'NA' > par3 = 'NA' > par2 = 'Croston' > par1 = 'Input box' > if(par3!='NA') par3 <- as.numeric(par3) else par3 <- NA > if(par4!='NA') par4 <- as.numeric(par4) else par4 <- NA > par6 <- as.numeric(par6) #Seasonal Period > par9 <- as.numeric(par9) #Forecast Horizon > par10 <- as.numeric(par10) #Alpha > library(forecast) Loading required package: tseries Loading required package: quadprog Loading required package: zoo Attaching package: 'zoo' The following object(s) are masked from package:base : as.Date.numeric This is forecast 2.03 > if (par1 == 'CSV') { + xarr <- read.csv(file=paste('tmp/',par7,'.csv',sep=''),header=T) + numseries <- length(xarr[1,])-1 + n <- length(xarr[,1]) + nmh <- n - par9 + nmhp1 <- nmh + 1 + rarr <- array(NA,dim=c(n,numseries)) + farr <- array(NA,dim=c(n,numseries)) + parr <- array(NA,dim=c(numseries,8)) + colnames(parr) = list('ME','RMSE','MAE','MPE','MAPE','MASE','ACF1','TheilU') + for(i in 1:numseries) { + sindex <- i+1 + x <- xarr[,sindex] + if(par2=='Croston') { + if (i==1) m <- croston(x,alpha=par10) + if (i==1) mydemand <- m$model$demand[] + fit <- croston(x[1:nmh],h=par9,alpha=par10) + } + if(par2=='ARIMA') { + m <- auto.arima(ts(x,freq=par6),d=par3,D=par4) + mydemand <- forecast(m) + fit <- auto.arima(ts(x[1:nmh],freq=par6),d=par3,D=par4) + } + if(par2=='ETS') { + m <- ets(ts(x,freq=par6),model=par5) + mydemand <- forecast(m) + fit <- ets(ts(x[1:nmh],freq=par6),model=par5) + } + try(rarr[,i] <- mydemand$resid,silent=T) + try(farr[,i] <- mydemand$mean,silent=T) + if (par2!='Croston') parr[i,] <- accuracy(forecast(fit,par9),x[nmhp1:n]) + if (par2=='Croston') parr[i,] <- accuracy(fit,x[nmhp1:n]) + } + write.csv(farr,file=paste('tmp/',par8,'_f.csv',sep='')) + write.csv(rarr,file=paste('tmp/',par8,'_r.csv',sep='')) + write.csv(parr,file=paste('tmp/',par8,'_p.csv',sep='')) + } > if (par1 == 'Input box') { + numseries <- 1 + n <- length(x) + if(par2=='Croston') { + m <- croston(x) + mydemand <- m$model$demand[] + } + if(par2=='ARIMA') { + m <- auto.arima(ts(x,freq=par6),d=par3,D=par4) + mydemand <- forecast(m) + } + if(par2=='ETS') { + m <- ets(ts(x,freq=par6),model=par5) + mydemand <- forecast(m) + } + summary(m) + } Forecast method: Croston's method Model Information: $demand Point Forecast Lo 80 Hi 80 Lo 95 Hi 95 18 36.90144 16.01576 57.78713 4.959547 68.84334 19 36.90144 15.91159 57.89130 4.800235 69.00265 20 36.90144 15.80794 57.99495 4.641709 69.16118 21 36.90144 15.70479 58.09810 4.483959 69.31893 22 36.90144 15.60214 58.20074 4.326972 69.47591 23 36.90144 15.49999 58.30290 4.170739 69.63215 24 36.90144 15.39832 58.40457 4.015248 69.78764 25 36.90144 15.29712 58.50576 3.860489 69.94240 26 36.90144 15.19640 58.60648 3.706451 70.09644 27 36.90144 15.09615 58.70674 3.553124 70.24976 $period Point Forecast Lo 80 Hi 80 Lo 95 Hi 95 18 1 1 1 1 1 19 1 1 1 1 1 20 1 1 1 1 1 21 1 1 1 1 1 22 1 1 1 1 1 23 1 1 1 1 1 24 1 1 1 1 1 25 1 1 1 1 1 26 1 1 1 1 1 27 1 1 1 1 1 In-sample error measures: ME RMSE MAE MPE MAPE MASE -3.2699313 16.1149252 13.5267259 -51.8842213 73.0748679 0.7559461 Forecasts: Point Forecast 18 36.90144 19 36.90144 20 36.90144 21 36.90144 22 36.90144 23 36.90144 24 36.90144 25 36.90144 26 36.90144 27 36.90144 > postscript(file="/var/www/html/rcomp/tmp/11pji1273760221.ps",horizontal=F,pagecentre=F,paper="special",width=8.3333333333333,height=5.5555555555556) > op <- par(mfrow=c(2,1)) > if (par2=='Croston') plot(m) > if ((par2=='ARIMA') | par2=='ETS') plot(forecast(m)) > plot(mydemand$resid,type='l',main='Residuals', ylab='residual value', xlab='time') > par(op) > dev.off() null device 1 > postscript(file="/var/www/html/rcomp/tmp/2bgi31273760221.ps",horizontal=F,pagecentre=F,paper="special",width=8.3333333333333,height=5.5555555555556) > op <- par(mfrow=c(2,2)) > acf(mydemand$resid, lag.max=n/3, main='Residual ACF', ylab='autocorrelation', xlab='time lag') > pacf(mydemand$resid,lag.max=n/3, main='Residual PACF', ylab='partial autocorrelation', xlab='time lag') > cpgram(mydemand$resid, main='Cumulative Periodogram of Residuals') > qqnorm(mydemand$resid); qqline(mydemand$resid, col=2) > par(op) > dev.off() null device 1 > > #Note: the /var/www/html/rcomp/createtable file can be downloaded at http://www.wessa.net/cretab > load(file="/var/www/html/rcomp/createtable") > > a<-table.start() > a<-table.row.start(a) > a<-table.element(a,'Demand Forecast',6,TRUE) > a<-table.row.end(a) > a<-table.row.start(a) > a<-table.element(a,'Point',header=TRUE) > a<-table.element(a,'Forecast',header=TRUE) > a<-table.element(a,'95% LB',header=TRUE) > a<-table.element(a,'80% LB',header=TRUE) > a<-table.element(a,'80% UB',header=TRUE) > a<-table.element(a,'95% UB',header=TRUE) > a<-table.row.end(a) > for (i in 1:length(mydemand$mean)) { + a<-table.row.start(a) + a<-table.element(a,i+n,header=TRUE) + a<-table.element(a,as.numeric(mydemand$mean[i])) + a<-table.element(a,as.numeric(mydemand$lower[i,2])) + a<-table.element(a,as.numeric(mydemand$lower[i,1])) + a<-table.element(a,as.numeric(mydemand$upper[i,1])) + a<-table.element(a,as.numeric(mydemand$upper[i,2])) + a<-table.row.end(a) + } > a<-table.end(a) > table.save(a,file="/var/www/html/rcomp/tmp/378gu1273760221.tab") > a<-table.start() > a<-table.row.start(a) > a<-table.element(a,'Actuals and Interpolation',3,TRUE) > a<-table.row.end(a) > a<-table.row.start(a) > a<-table.element(a,'Time',header=TRUE) > a<-table.element(a,'Actual',header=TRUE) > a<-table.element(a,'Forecast',header=TRUE) > a<-table.row.end(a) > for (i in 1:n) { + a<-table.row.start(a) + a<-table.element(a,i,header=TRUE) + a<-table.element(a,x[i]) + a<-table.element(a,x[i] - as.numeric(m$resid[i])) + a<-table.row.end(a) + } > a<-table.end(a) > table.save(a,file="/var/www/html/rcomp/tmp/4ihfw1273760221.tab") > a<-table.start() > a<-table.row.start(a) > a<-table.element(a,'What is next?',1,TRUE) > a<-table.row.end(a) > a<-table.row.start(a) > a<-table.element(a,hyperlink(paste('http://www.wessa.net/Patrick.Wessa/rwasp_demand_forecasting_simulate.wasp',sep=''),'Simulate Time Series','',target='')) > a<-table.row.end(a) > a<-table.row.start(a) > a<-table.element(a,hyperlink(paste('http://www.wessa.net/Patrick.Wessa/rwasp_demand_forecasting_croston.wasp',sep=''),'Generate Forecasts','',target='')) > a<-table.row.end(a) > a<-table.row.start(a) > a<-table.element(a,hyperlink(paste('http://www.wessa.net/Patrick.Wessa/rwasp_demand_forecasting_analysis.wasp',sep=''),'Forecast Analysis','',target='')) > a<-table.row.end(a) > a<-table.end(a) > table.save(a,file="/var/www/html/rcomp/tmp/5b9xz1273760221.tab") > try(system("convert tmp/11pji1273760221.ps tmp/11pji1273760221.png",intern=TRUE)) character(0) > try(system("convert tmp/2bgi31273760221.ps tmp/2bgi31273760221.png",intern=TRUE)) character(0) > > #-SERVER-wessa.org > > > > proc.time() user system elapsed 1.389 0.343 1.580