R语言forecast包

Package‘forecast’December22,2014Version5.7Date2014-12-17TitleForecastingfunctionsfortimeseriesandlinearmodelsDescriptionMethodsandtoolsfordisplayingandanalysingunivariatetimeseriesforecastsincludingexponentialsmoothingviastatespacemodelsandautomaticARIMAmodelling.DependsR(=3.0.2),stats,graphics,zoo,timeDateImportstseries,fracdiff,Rcpp(=0.11.0),nnet,colorspace,parallelSuggeststestthat,fppLinkingToRcpp(=0.11.0),RcppArmadillo(=0.2.35)[e-mail]GeorgeAthanasopoulos,SlavaRazbash,DrewSchmidt,ZhenyuZhou,YousafKhan,ChristophBergmeir,[e-mail](=2)URL:21:28Rtopicsdocumented:accuracy...........................................3Acf.............................................4arfima............................................512Rtopicsdocumented:Arima............................................7arima.errors.........................................9arimaorder..........................................10auto.arima..........................................11bats.............................................13bizdays...........................................14BoxCox...........................................15BoxCox.lambda.......................................16croston............................................17CV..............................................19dm.test............................................19dshw.............................................21easter............................................23ets..............................................23findfrequency........................................26fitted.Arima.........................................27forecast...........................................27forecast.Arima.......................................29forecast.bats.........................................31forecast.ets.........................................32forecast.HoltWinters....................................34forecast.lm.........................................35forecast.stl..........................................37forecast.StructTS......................................39gas..............................................40getResponse.........................................41gold.............................................42logLik.ets..........................................42ma..............................................43meanf............................................44monthdays..........................................45msts.............................................46na.interp...........................................47naive.............................................48ndiffs............................................49nnetar............................................51plot.bats...........................................52plot.ets............................................53plot.forecast.........................................54rwf..............................................55seasadj............................................57seasonaldummy.......................................58seasonplot..........................................59ses..............................................60simulate.ets.........................................62sindexf............................................64splinef............................................65subset.ts...........................................66accuracy3taylor............................................67tbats.............................................68tbats.components......................................69thetaf............................................70tsclean............................................72tsdisplay...........................................73tslm.............................................74tsoutliers...........................................75wineind...........................................76woolyrnq..........................................76Index77accuracyAccuracymeasuresforforecastmodelDescriptionReturnsrangeofsummarymeasuresoftheforecastaccuracy.Ifxisprovided,thefunctionmeasuresout-of-sample(testset)forecastaccuracybasedonx-f.Ifxisnotprovided,thefunctiononlyproducesin-sample(trainingset)accuracymeasuresoftheforecastsbasedonf[x]-fitted(f).AllmeasuresaredefinedanddiscussedinHyndmanandKoehler(2006).Usageaccuracy(f,x,test=NULL,d=NULL,D=NULL)ArgumentsfAnobjectofclassforecast,oranumericalvectorcontainingforecasts.ItwillalsoworkwithArima,etsandlmobjectsifxisomitted–inwhichcasein-sampleaccuracymeasuresarereturned.xAnoptionalnumericalvectorcontainingactualvaluesofthesamelengthasobject,oratimeseriesoverlappingwiththetimesoff.testIndicatorofwhichelementsofxandftotest.IftestisNULL,allelementsareused.Otherwisetestisanumericvectorcontainingtheindicesoftheelementstouseinthetest.dAnintegerindicatingthenumberoflag-1diff