Beruflich Dokumente
Kultur Dokumente
Chapter
1: STATISTICA: General Overview 1
2: Step-by-Step Examples 9
Analytics 11
Data Management 72
Enterprise Installations 98
3: User Interface 125
4: Output from Analyses 145
5: STATISTICA Documents 167
6: Graphs 187
7: Customizing STATISTICA 211
8: STATISTICA Visual Basic 219
9: STATISTICA Query 241
10: STATISTICA and .NET 247
Appendixes
A: Getting More Help 255
B: STATISTICA Enterprise Server 261
C: STATISTICA Family of Products 273
QuickReference
STATISTICA:
A GENERAL OVERVIEW
OF FEATURES
1
1
CHAPTER
CHAPTER
2
2
CHAPTER2: ENTERPRISE EXAMPLES
YoucanalsodoubleclickoneitherSTATIST.exeinWindowsExplorerortheiconof
anySTATISTICAfile,e.g.,aspreadsheet,tostarttheprogram.
WhenyoustartSTATISTICAforthefirsttime,theUserInterfacedialogisdisplayed,
whereyoucanchoosetousetheribbonbarortheclassicdropdownmenus.All
examplesinthismanualusetheribbonbar.
CHAPTER
2
2
Chapter2: StepbyStep Examples
Notethatitiseasytoswitchbetweentheribbonbarandtheclassicmenusatany
time.Whentheribbonbarisdisplayed,clickthemenuicon ontheQuickAccess
toolbar(locatedintheupperleftcorneroftheribbonbar)todisplaytheclassic
menus.Whentheclassicmenusaredisplayed,selectRibbonBarfromtheView
menutodisplaytheribbonbar.
Tocreatemorespaceintheapplicationwindow,youcanminimizetheribbonbar.
Eitherdoubleclickontheselectedtabheader,orrightclickontherightsideofthe
rowoftabsandfromtheshortcutmenu,selectMinimizetheRibbon.
AfteryouclickOKintheUserInterfacedialog,theWelcometo
STATISTICAdialogisdisplayed,whichcontainsoptionsthatare
usefultoaccesscommonfunctionsinSTATISTICA.
Ifyouprefer,youcanselecttheDontshowthisdialogagain
checkboxlocatednearthebottomofthedialog,andthisdialog
willnotbedisplayedwhenyoustartSTATISTICA.Dependingon
theversionofSTATISTICAyouhave,theremaybeotherdialogs
displayedaswell.
Customization of STATISTICA.Practicallyallaspectsofthe
behaviorandappearanceofSTATISTICA(evenmanyelementary
featuresillustratedinthisexample,suchaswhereoutputis
directed)canbepermanentlycustomizedtomatchyour
preferences.Forexample,eventhefirststep(openingSTATISTICA)canbe
customized;youcanchangethedefaultfullscreenopeningmode,theappearance
Chapter2:StepbyStep Examples
AllthecommandsontheribbonbarandclassicmenusaredescribedinSTATISTICA
Help;pointto(highlight)acommand,andpressF1onyourkeyboardtodisplaythe
respectiveHelptopic.
Variable specifications.Thevariable(column)headersinthespreadsheet
containthevariablenames.DoubleclickonthefirstvariableheaderGENDER
todisplayitsVariablespecificationsdialog.
Chapter2: StepbyStep Examples
Spreadsheet formulas.Usingtheoptionsinthisdialog,youcanchangethe
variablenameand/orformat,enteraformulatorecalculatethevaluesofthe
variable,etc.IftheentryintheLongname(labelorformulawithFunctions)box
startswithanequalsign(=),STATISTICAinterpretsitasaformula[acommentcan
followafterasemicolon(;)].Forexample,ifyouenterintotheLongnamebox
(ofvariableone)=(v2+v3+v4)/3or=mean(v2:v4),thecurrentvaluesofthat
variablewillbereplacedbytheaverageofvariablestwothroughfour,separately
foreachcase(row)ofthespreadsheet.
Specificationsofallvariablescanalsobereviewedandeditedtogetherina
combinedVariableSpecificationsEditordialog,accessedbyclickingtheAll
SpecsbuttonintheVariablespecificationsdialog.
ThereareanumberofwaystooutputtotheWeb,dependingontheversionof
STATISTICAyouhave.Thesemeansforoutputcanbeusedinmanycombinations
(e.g.,aworkbookandreportsimultaneously),andeachoutputchannelcanbe
customizedinavarietyofways.Also,alloutputobjects(spreadsheetsandgraphs)
cancontainotherembeddedandlinkedobjectsanddocuments,soSTATISTICA
outputcanbehierarchicallyorganizedinavarietyofways.
Calculating a correlation matrix.Now,letscomputeacorrelationmatrixforthe
variablesintheAdstudy.stadatafile.TodisplaytheBasicStatisticsandTables
StartupPanel,selecttheStatisticstab,andintheBasegroup,clickBasicStatistics,
Chapter2:StepbyStep Examples
in
thelowerleftcornerofthescreen.
Atthispoint,ensurethatablock(agroupofselectedcells)isnotselectedinthe
spreadsheet.Todeselectablock,clickinanycellinthespreadsheet.Ifablockis
selected,STATISTICAassumesthatthevariablescorrespondingtotheblockare
intentionallypreselectedfortheanalysis,andwhenyoulaterclicktheOKor
Summarybuttontoproducetheanalysisresults,insteadofpromptingyouto
selectvariables,STATISTICAwillautomaticallyproducethecorrelationsforthe
selectedblockvariables.
IntheBasicStatisticsandTablesStartupPanel(showninthenextillustration),
selectCorrelationmatricesandclicktheOKbutton(ordoubleclickCorrelation
matrices)todisplaytheProductMomentandPartialCorrelationsdialog.
Chapter2: StepbyStep Examples
Thevariableselectiondialogsupportsvariouswaysofselectingvariables(including
thestandardWindowsSHIFT+clickandCTRL+clickconventionstoselectrangesand
discontinuouslistsofvariables).
Youcanalsousevariousshortcutsandoptionsinthevariableselectiondialogto
reviewthecontentsofthedatafile.Forexample,youcanspreadthevariablelist
Chapter2: StepbyStep Examples
Forthisexample,selectvariables1through10inthevariableselectiondialog.
ClicktheOKbutton.Amessagewillbedisplayedinformingyouthattherearetext
variablesselected.ClicktheContinuewithcurrentselectionbuttontoreturnto
theProductMomentandPartialCorrelationsdialog.Next,clicktheSummary
buttontogenerateacorrelationmatrixfortheselectedvariables.
NotethatinsteadofclickingtheSummarybutton,youcouldhaveclickedthe
Summary:CorrelationsbuttonontheQuicktaborontheAdvancedtabwiththe
Chapter2:StepbyStep Examples
Thesegraphsnotonlyshowthescatterplotofpointsforeachcorrelation,butalso
thedistributions(histograms)foreachvariable,aswellastherespective
correlationcoefficientandregressionequation.
STATISTICAincorporatesmanysuchdisplaystosummarizebasicdescriptive
statistics,correlations,theresultsofGageorProcesscapabilitystudies,orother
typesofdataanalyses.
Results spreadsheets (multimedia tables).Inadditiontostoringdata,
spreadsheetsareusedinSTATISTICAtodisplaymostofthenumericoutput.Note
thatspreadsheetsoffermanydisplayfeaturesandoptions,andinthisexample,
significantcorrelationsaremarkedwithadifferentformattohelpdistinguish
them;bydefault,thecolorisred(intheCorrelationsspreadsheet,seethecell
adjacenttoMEASURE07underGENDER).Spreadsheetscanholdanywherefroma
shortlinetogigabytesofoutput,andtheyofferavarietyofoptionstofacilitate
reviewingtheresultsandvisualizingtheminpredefinedandcustomdefined
Chapter2: StepbyStep Examples
Tochangeoutputoptionsforallanalyses,usethe(global)OutputManager(the
OutputManageroptionspaneoftheOptionsdialog,accessiblebyselectingthe
HometabandclickingOptionsintheToolsgroup),orselecttheUseglobalOutput
settings(changesherewillaffecttheglobalsettings)optionbuttoninthe
Analysis/GraphOutputManagerdialog.
Aswithallworkbooks,individualdocuments(e.g.,spreadsheetsorgraphs)or
groupsofdocumentscanbeprinted,extracted,copied,anddeletedfroman
analysisworkbook.SeetheoverviewofWorkbooksonpage169formoredetails;
seealsotheElectronicManual(STATISTICAHelp).
Copy vs. Copy with Headers.Contentsofspreadsheetscanbecopiedtothe
ClipboardbypressingCTRL+C(whichcopiesonlythecontentsoftheselectedblock).
Tocopytheblockalongwithitsrespectivevariableandcasenames,selecttheEdit
tab,andintheClipboard/Datagroup,clicktheCopyarrowandselectCopywith
Headersfromthedropdownmenu.Whenspreadsheetsarepastedintoaword
processordocument,theywillbeactive(inplaceeditable)STATISTICAobjects,
Chapter2: StepbyStep Examples
Chapter2:StepbyStep Examples
button
in
anyanalysisorgraphspecificationdialog,andselectOutput(forlocalchanges).
IntheOutputManageroptionspaneoftheOptionsdialogorinthe
Analysis/GraphOutputManagerdialog,clicktheReportOutputarrow.Fromthe
dropdownmenu,selecteitherSendtoMultipleReports(oneforeach
Analysis/Graph),SingleReport(commonforallAnalyses/graphs),or[SelectFile]
(whichwilldisplaytheOpendialogwhereyoucanselectanalreadyestablished
report).
IntheOutputManager,youcanalsospecifytheamountofsupplementary
informationtobeincludedwiththespreadsheetresults.UsetheSupplementary
detailoptiontospecifyeitherBrief(includesonlytheselectedspreadsheetsand
graphs),Medium(includestheselectedspreadsheetsandgraphsaswellasthe
currentdatafilename,informationoncaseselectionconditionsandcaseweights
ifanywerespecified,alistofallvariablesselectedforeachanalysis,andthe
missingdatavaluesforeachvariable),Long[includesallinformationfromthe
Mediumformatandthelongvariablelabels(e.g.,formulas),reservingonelineof
output(ormore)foreachvariable],orComprehensive(includesallinformation
includedintheLongreportformataswellasacompletelistofallofthetextlabels
foreachselectedvariable).
Interpretation of the results STATISTICA Electronic Manual (Help) and the
Electronic Statistics Textbook.Nowletsreturntotheexampleandthe
correlationmatrixthathasbeenproduced.
Chapter2:StepbyStep Examples
Toopenthetextbook,selecttheHelptab,andintheHelpgroup,clickElectronic
StatsTextbook.
Also,manytopicsinSTATISTICAHelpcontainalinktothetextbook.
Clickthelinkintheupperrightcornerofthetopictodisplaytherespectivepagein
theElectronicTextbook.
Chapter2: StepbyStep Examples
Thespecifiedgraphwillbedisplayed.
Aswecanlearnfromthegraph,therearenounusualpatternsofdata,thus,there
isnoreasontobeconcernedaboutoutliers(seetheshortdiscussionofoutlierson
page28;seealsothetopiconoutliersintheElectronicManual).
Graph customization.Notethatnow,whenthefocusisonthegraphwindow,the
Edittabcontainsdifferentoptionsthanitdidforthespreadsheets.
Itcontainsavarietyofgraphcustomizationanddrawingtools.Manyofthese
optionsarealsoavailablefromshortcutmenusaccessedbyrightclickingon
specificpartsofthegraph.Notethattheoptionsonshortcutmenusare
Chapter2: StepbyStep Examples
Formoreinformationongraphcustomization,seepage190andtheElectronic
Manual.
Nowletsreturntothespreadsheet.
Split scrolling in spreadsheets.Spreadsheetscanbesplitintouptofoursections
(panes)bydraggingthesplitbox(thesmallrectangleatthetopofthevertical
scrollbarortotheleftofthehorizontalscrollbar).Thisisusefulifyouhavealarge
amountofinformationandyouwanttoreviewresultsfromdifferentpartsofthe
spreadsheet.Whenyoumovethemousepointertothesplitbox,themouse
pointerchangesto or .Now,topositionthesplit,dragittothedesired
position.
Chapter2:StepbyStep Examples
Notethatverticallysplitpanesscrolltogetherwhenyouscrollhorizontally;
horizontallysplitpanesscrolltogetherwhenyouscrollvertically.Forinformation
abouthighlightingblocksofdataacrosssplitpanesandaboutvariablespeed
highlightingofblocksofdata,seeHowcanIexpandablockinthespreadsheet
outsidethecurrentscreen?intheElectronicManual.
Drag-and-drop.STATISTICAsupportsthecompletesetofstandardspreadsheet
(MicrosoftExcelstyle)draganddropfacilities.Forexample,inordertomovea
block,pointtotheborderoftheselection(themousepointerchangestoan
arrow)anddragittothenewlocation.
Tocopyablockofdata,pointtotheborderoftheselection(themousepointer
changestoanarrow),anddragtheselectiontoanewlocationwhilepressingthe
CTRLkey.Notethatwhenyouaredraggingtheselection,aplussign(+)isdisplayed
nexttothemousepointertoindicateyouarecopyingthetextratherthanmoving
it(seethenextimage).
Chapter2: StepbyStep Examples
Toinsertablockbetweencolumnsorrows,pointtotheborderoftheselection
(themousepointerchangestoanarrow)andthendragtheselectionwhile
pressingtheSHIFTkey.
Ifyoupointbetweenrows,aninsertionbarisdisplayedbetweentherows,and
whenyoureleasethemousebutton,theblockisinsertedbetweenthosetworows
[creatingnewcase(s)].Ifyoupointbetweencolumns,aninsertionbarisdisplayed
betweenthecolumns,andwhenyoureleasethemousebutton,theblockis
insertedbetweenthosetwocolumns[creatingnewvariable(s)].
NotethatifyoualsopresstheCTRLkeywhileyouaredraggingtheselection,the
blockwillbecopiedandinsertedinsteadofmovedandinserted;apluswillappear
nexttothemousepointer(asshowninthenextillustration).
Additionally,aseriesofvalueswithinablockcanbeextrapolated(AutoFilled)by
draggingtheFillHandle(thesmall,solidsquarelocatedonthelowerrightcorner
oftheblockborder).
Chapter2:StepbyStep Examples
Thisdialogisusedtospecifyverysimpleanalyses(e.g.,viaOnewayANOVA
designswithonlyonebetweengroupfactor)andmorecomplexanalyses(e.g.,via
RepeatedmeasuresANOVAdesignswithbetweengroupfactorsandawithin
subjectfactor).
Design.SelectRepeatedmeasuresANOVAastheTypeofanalysisandQuick
specsdialogastheSpecificationmethod,andthenclicktheOKbuttoninthe
GeneralANOVA/MANOVAStartupPaneltodisplaytheANOVA/MANOVA
RepeatedMeasuresANOVAdialog.
ThenclicktheOKbuttontoreturntotheANOVA/MANOVARepeatedMeasures
ANOVAdialog.
The repeated measures design.Thedesignoftheexperimentthatwearegoing
toanalyzecanbesummarizedasfollows:
Between-Group Between-Group Repeated Measure Factor: Response
Factor #1:
Gender
Factor #2:
Advert
Level #1:
Measure01
Level #2:
Measure02
Level #3:
Measure03
Subject 1 Male Pepsi 9 1 6
Subject 2 Male Coke 6 7 1
Subject 3 Female Coke 9 8 2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Specifying a repeated measures factor.Theminimumnecessaryselectionsare
nowcomplete,and,ifwedidnotwanttoselecttherepeatedmeasuresfactor,we
wouldbereadytoclicktheOKbuttonandseetheresultsoftheanalysis.However,
forourexample,weneedtospecifythatthethreedependentvariableswehave
selectedbeinterpretedasthreelevelsofarepeatedmeasures(withinsubject)
factor.Unlesswedoso,STATISTICAassumesthatthosearethreedifferent
dependentvariablesandrunsaMANOVA(i.e.,MultivariateANOVA).
Chapter2: StepbyStep Examples
NotethatSTATISTICAhassuggestedtheselectionofonerepeatedmeasuresfactor
with3levels(defaultnameR1).Youcanspecifyonlyonewithinsubject(repeated
measures)factorviathisdialog.Tospecifymultiplewithinsubjectfactors,usethe
GeneralLinearModelsmodule(availableintheoptionalAdvanced
Linear/NonlinearModelspackage).PresstheF1keyonyourkeyboardwhilethe
Specifywithinsubjectsfactordialogisdisplayed(orclickthe
buttoninthe
upperrightcornerofthedialog)todisplaytheElectronicManualtopicthat
describesalloptionsinthisdialogandcontainslinkstocomprehensivediscussions
ofrepeatedmeasuresandexamplesofdesigns.
Forthisexample,editthenameforthefactor:intheFactorNamebox,changethe
defaultR1toRESPONSE,andclicktheOKbuttontoexitthedialog.
Codes (defining the levels) for between-group factors.Youdonotneedto
manuallyspecifycodesforbetweengroupfactors[i.e.,thereisnoneedtoinstruct
STATISTICAthatvariableGenderhastwolevels:1and2(orMaleandFemale)]
unlessyouwanttopreventSTATISTICAfromusing,bydefault,allcodes
encounteredintheselectedgroupingvariablesinthedatafile.Toentersuch
customcodeselection,clicktheFactorcodesbuttontoaccesstheSelectcodesfor
indep.vars(factors)dialog.
Beforeyoumakeyourselections,youcanusetheoptionsinthisdialogtoreview
valuesofindividualvariablesbyclickingtheZoombutton,scanthefile,andfillin
thecodesfields(e.g.,GenderandAdvert)foranindividualvariableorallvariables,
etc.Fornow,clicktheOKbuttonintheSelectcodesforindep.vars(factors)
Chapter2:StepbyStep Examples
andclosesthedialog.
Performing the analysis.ClicktheOKbuttonintheANOVA/MANOVARepeated
MeasuresANOVAdialog.TheanalysisisperformedandtheANOVAResultsdialog
isdisplayed,whichcontainsvariousoutputspreadsheetsandgraphsoptions.
Thisdialogcontainsseveraltabsthatenableyoutoquicklylocatethedesired
resultsoptions.Forexample,ifyouwanttoperformplannedcomparisons,select
theCompstab.Toviewresidualstatistics,selecttheResidstab.Forthisexample,
wewillonlyusetheresultsoptionsavailableontheQuicktab.
Reviewing ANOVA results.LetsstartbylookingattheANOVAsummaryofall
effectstablebyclickingtheAlleffectsbutton(theonewiththeSUMMicon ).
Theonlyeffect(ignoringtheIntercept)inthisanalysisthatisstatisticallysignificant
(p=.007)istheRESPONSEeffect.Thisresultmaybecausedbymanypossible
patternsofmeansoftheRESPONSEeffect(formoreinformation,consultthe
Chapter2: StepbyStep Examples
Thisdialogcontainsasummarytableofalleffects(withmostoftheinformation
youhaveseeninthealleffectsspreadsheet)andisusedtoreviewindividual
effectsfromthattableintheformoftheplotsoftherespectivemeans(or,
optionally,spreadsheetsoftherespectivemeanvalues).
Plot of means for a main effect.IntheTableofAllEffectsdialog,doubleclickon
thesignificantmaineffectRESPONSE(theonemarkedwithanasteriskinthep
column)toproducetherespectiveplot.
Thegraphindicatesthatthereisacleardecreasingtrend;themeansforthe
consecutivethreequestionsaregraduallylower.Eventhoughthereareno
significantinteractionsinthisdesign(seethediscussionoftheTableofalleffects,
Chapter2:StepbyStep Examples
ClicktheOKbuttontoacceptthedefaultarrangementandproducetheplotof
means.
Asyoucansee,thispatternofmeans(splitbythelevelsofthebetweengroup
factors)doesnotindicateanysalientdeviationsfromtheoverallpatternrevealed
inthefirstplot(forthemaineffect,RESPONSE).Nowyoucancontinueto
interactivelyexamineothereffectsrunposthoccomparisons,planned
comparisons,extendeddiagnostics,etc.tofurtherexploretheresults.
Interactive data analysis in STATISTICA.Thisexampleillustratesthewayin
whichSTATISTICAsupportsinteractivedataanalysis.Youarenotforcedtospecify
Chapter2: StepbyStep Examples
ClicktheNewbuttontodisplaytheNewBundledialog,
enterthenameProductionintheBundlenamefield,andclicktheOKbutton.The
Selectvariablesforbundledialogisdisplayed,whichcontainsallthevariablesin
theEnginePerformance.stadataset.
Forouranalyses,weneedtoselectthevariablesInput01Input05,Input20,
Input30Input35,andInput70.Youcanselectthesevariablesusingthestandard
WindowsSHIFT+clickandCTRL+clickconventionstoselectrangesanddiscontinuous
listsofitems,respectively.
ClicktheOKbuttontoclosetheSelectvariablesforbundledialogandreturnto
theVariableBundleManager.
Chapter2: StepbyStep Examples
Theleftpaneofthisdialogdisplaysthenamesofallbundlesthathavebeen
definedforthisspreadsheet(youcancreatenumerousbundlesineach
spreadsheetifneeded).Therightpanedisplaysthecontentsofthebundlethatis
currentlyselectedintheleftpane.Ifbothofthesepanesareempty,nobundles
havebeencreatedforthisspreadsheet.
YoucanmakechangestoabundlebyclickingtheEditbutton,discardabundleby
clickingtheDeletebutton,changethetitleofabundlebyclickingtheRename
button,andproduceaspreadsheetcontaininginformationaboutthebundlesfor
theactivedataspreadsheetbyclickingtheOutputtoSpreadsheetbutton.
Forthisexample,clicktheOKbuttontoacceptthebundlewecreatedandclose
theVariableBundleManagerdialog.
Then,selecttheStatisticstab,andintheBasegroup,clickMultipleRegressionto
displaytheMultipleLinearRegressionStartupPanel.OntheQuicktab,clickthe
Variablesbuttontodisplaythevariablespecificationdialog.
Bundlesaredisplayedinbracketsandlisted(inalphabeticalorder)atthetopof
thevariablelist.IntheIndependentvariablelist,selecttheProductionbundleto
specifywithoneclickofthemousebuttonInput01Input05,Input20,Input30
Input35,andInput70astheindependentvariablesfortheanalysis.
Chapter2:StepbyStep Examples
Additionally,youcanviewthelistofvariables(byname)byclickingthe[Bundles]
buttoninthevariablespecificationdialog.ThisdisplaystheVariableBundles
Manager.
Notethatbundlesaredefinedforasinglespreadsheet,andtheyareonlyusedfor
variableselection.Hence,theyareneverlistedinreportsorotheroutput.
Asyoucanseewiththisexample,youwillsaveconsiderabletimebyselectinga
bundleratherthanlookingforthecorrectvariablestochooseinalargedataset.
Example 4: By-Group Analyses
STATISTICAoffersapowerfuloptiontoturneverystatisticalorgraphicsanalysis
intoananalysisbygroup.Whenreviewingresultsintheresultsdialogofpractically
anyanalysis,orusingthegraphsoptions,youcanselectoneormoregrouping
variables,andthencreateresults1)forallcasesinthedatacombined,and/or
2)brokendownbyeachcombinationofuniquevaluesinthegroupingvariables.
Chapter2: StepbyStep Examples
Thisisaverypowerfultoolforinteractiveandexploratorydataanalysis,allowing
youtoreviewquicklywhetheranypatternsorspecificresultsholdinall
subgroups,samples,orstratainyourdata.
Forexample,youmaybeperformingamultipleregressionanalysisanddecideto
review,withoutexitingthecurrentdialog,theresultsbrokendownbyGenderand
anothergroupingvariableinyourdata.Afterselecting(enabling)thisoption(by
clickingthe ByGroupbutton),everytimeyouclickanyoftheresultsbuttons
(e.g.,tocreateasummaryresultsspreadsheetorgraph),allresultsarecomputed
notonlyforallgroups(optionally),butalsoforeachuniquecombinationof
groupingvariablesthatwerespecified(e.g.,byGenderandanothergrouping
variable).
TheresultsoftheByGroupanalysiscanbeplacedeitherinthedefaultresults
workbookintotheirownfolder,labeledwiththerespectivebygroupcondition
(e.g.,Gender=Female;Time=After1),orintothesamefolderwithallotherresults.
Chapter2:StepbyStep Examples
Forexample,youcouldcreatemultiplelineplotstodescribeamultivariatebatch
process,creatingaseparategraph(trajectories)foreachbatch.
Exploring Experimental Data Using
the By Group Option
ThisexampleisbasedonthedatafileTomatoes.sta,whichisoneoftheexample
datafilesdescribedingreaterdetailintheExperimentalDesignsectionofthe
STATISTICAElectronicManual(seetheexampleDesigningandAnalyzinga2
3
3
2
Experiment).ConnorandYoung(inMcLeanandAnderson,1984)reportan
experiment(takenfromYoudenandZimmerman,1936)onvariousmethodsof
producingtomatoplantseedlingspriortotransplantinginthefield.
StartbyopeningtheexampleTomatoes.stadataset.SelecttheHometab.Inthe
Filegroup,clicktheOpenarrowandselectOpenExamplesfromthedropdown
menutodisplaytheOpenaSTATISTICADataFiledialog.Doubleclickonthe
Datasetsfolder,andthenselectandopentheSTATISTICAdatasetTomatoes.sta.
Chapter2: StepbyStep Examples
Shownhereareafewrows(cases)ofthatdatafile.Youcanrefertothe
ExperimentalDesignElectronicHelpexampletopicforacompleteanalysisofthese
data.
Exploring Patterns by Variety
Thisexampleillustratesatypicalworkflowasitoftenappliestotheanalysisof
discreteorbatchmanufacturingdata,i.e.,thegoaloftheanalysisistoverify
(graphicallyoranalytically)thatsomepatternsordistributionsequallyapplytoall
samples,parts,orbatches.
WewillexploretheeffectofProductionMethod,SoilCondition,andPotsizeon
yield(Pounds),andevaluatewhetheranypatternsholdforeachVarietyinthe
study.Insteadofperformingacompleteanalysisofvariance(asisdescribedinthe
ExperimentalDesignexampleoftheElectronicHelp),wewillusemostlygraphical
methodsandvisualinspection.
Specifying variability plots.SelecttheGraphstab.IntheMoregroup,click2D,
andfromthedropdownmenu,selectVariabilityPlotstodisplaytheVariability
Plotdialog.ClicktheVariablesbutton,andintheSelectVariablesforVariability
Plotdialog,selectPOUNDSastheDependentvariable,andSOILCONDITION,
POTSIZE,andPRODUCTIONMETHODfromtheGroupingvariablelist.
Chapter2:StepbyStep Examples
Finally,alsointheVariabilityPlotdialog,ensurethatPRODUCTIONMETHODis
selectedintheFactorslist,andselectthePutboxesaroundgroupscheckbox.
Specifying by grouping.WewanttocreatethevariabilityplotforPRODUCTION
METHOD,SOILCONDITION,andPOTSIZEforallvarietiesoftomatoescombined,
andbrokendownbyVARIETY(onegraphperVARIETY).ClicktheByGroupbutton
todisplaytheByGroupdialog.
ClicktheGroupingVariable(s)buttontodisplaytheSelectByVariablesdialog,
andspecifyVARIETYastheByGroupvariable.
Chapter2: StepbyStep Examples
NotethatyoucanspecifymorethanoneByGroupvariable,inwhichcaseall
subsequentanalyseswillbeperformedbrokendownbyeachuniquecombination
ofvaluesfoundintheByGroupvariables.
Reviewing the variability plots.NowclickOKtoclosetheSelectByVariables
dialog,andclickOKtoclosetheByGroupdialog.IntheVariabilityPlotdialog,
clickOKtocreatethegraphs.
NoticehowtheVariabilityPlotiscreated1)forAllGroups,and2)foreachVariety
(BonnyandMarglobe).
Ifyoureviewthesegraphscarefully,youwillseethattheProductionMethod
appearstomakelittledifference(intheobservedvaluesforPounds)for
Variety=Bonny,whileforVariety=Marglobe,theFibrePlmethodshowstheleast
variabilityinvalues,whicharegenerallyatthehigherendofthedistributionofall
valuesforvariablePounds.
Descriptive Statistics By Group
Letsnextusethedescriptivestatisticsoptionstofurtherexplorethis.Selectthe
Statisticstab.IntheBasegroup,clickBasicStatisticstodisplaytheBasicStatistics
Chapter2:StepbyStep Examples
Now,clickOKinthisdialogandclickOKintheByGroupdialog.IntheStatisticsby
GroupsResultsdialog,clickinsequence,1)theSummarybutton,2)theAnalysis
ofVariancebutton,and3)theInteractionplotsbutton.
Chapter2: StepbyStep Examples
Allresultsareplacedintotherespectivefolder,eithertheAllGroupsfolderorthe
Variety=BonnyorVariety=Marglobefolders.
Youcannowreviewtheseresultsforallgroupscombinedandbrokendownby
Variety;asyouwillsee,indeed,ProductionMethodappearstohaveaneffecton
yield(Pounds)forVariety=Marglobe,whilethereisnoindicationofsuchaneffect
forVariety=Bonny.
Summary
WithSTATISTICA,youcanperformadhocbygroupanalysesfromvirtuallyany
resultsdialog,reviewingresultsforallgroupscombinedorbrokendownbyoneor
moregroupingvariable.Thisverypowerfulfeatureforexploratorydataanalysis
canbeusedtocomparegroupsandverifyconsistencyofresultsacrossgroupsfor
anyanalysis.
Beforeconcludingthistopic,afewcommentsaboutthetechnicaldetailsregarding
theimplementationofthisfeaturemaybeuseful.Whenperformingbygroup
analyses,asillustratedinthisexample,theprogramwillactuallyrerunthe
analysesforeachgroup(andallgroups),leveragingtheSTATISTICAVisualBasic
macrocodethatisrecordedautomaticallyduringtheinteractiveanalyses,and
whichcanbesavedasmacrosasdescribedelsewhereinthismanual(seeChapter
8STATISTICAVisualBasic).Whenanalyzingverylargedataproblems(e.g.,very
largeunbalancedexperimentaldesignsorcomplexanalysesthatrequireiterated
computationsbeforeresultscanbedisplayed),theindividualanalysesmaytakeup
significantamountsofcomputingtime,inparticularwhentherearemanyunique
Chapter2:StepbyStep Examples
STATISTICAincorporatesmanysuchdisplaystosummarizebasicdescriptive
statistics,correlations,theresultsofgageorprocesscapabilitystudies,orother
typesofdataanalyses,asshowninthefollowingillustration.
Chapter2: StepbyStep Examples
ClicktheOKbuttonintheProcessAnalysisProceduresStartupPanel.Onthe
QuicktaboftheISO21747ProcessCapabilitySetupdialog,clicktheVariables
button.IntheSelectVariables(andoptionalgroupingvariable)dialog,select
variableSizeintheVariablesfortheanalyseslist,andSampleintheby...
(Time/Groupingvar.)list,andclickOK.
Chapter2: StepbyStep Examples
ClickOKtofinalizethischoiceandreturntotheISO21747ProcessCapability
Setupdialog.
Inthisdialog,therearenumerousotheroptionsavailabletomodifytherulesthat
areappliedtoselectthemostappropriatedistributionandtimedependent
distributionmodelforthedatasothattheappropriateprocesscapabilityindices
canbecomputed.Youcanclickthe buttonintheupperrightcornerofthe
dialogorpressF1todisplaytheSTATISTICAElectronicHelptopiccontainingspecific
detailsregardingalloptionsinthisdialog.Forexample,thedetailsregardingthe
(small)differencesintheDINandISOspecificationsarediscussedthere.
NowclicktheOKbuttonintheISO21747ProcessCapabilitySetupdialogto
performtheanalysesforvariableSize.
Reviewing results.IntheISO21747ProcessCapabilityResultsdialog,clickthe
Summarybuttontoreviewtheanalysissummarydisplay.
Chapter2:StepbyStep Examples
Asyoucansee,allrelevantdetails(asrecommendedinISO21747and/orDIN
55319)aresummarizedonasinglepage(document),whichcontainsall
informationnecessarytojudgetheprocessascapableornotcapable(or
questionable).
Attribute Gage Analysis
Foranotherexampleofthistypeofsummary(compound)displaysinSTATISTICA,
wewillperformanattributegageanalysis.
Ingeneral,anymeasurementsystemusedinmanufacturingmustbevalidatedto
ensurethattherespectivegagesmeasurethequalitycharacteristicofinterestwith
sufficientaccuracyandprecision.Often,agageofparticularimportanceistheone
thatdetermineswhetheramanufacturedpartisofsufficientqualitytobe
acceptedorrejected;inotherwords,thegagemeasuresasimpleaccept/reject
attribute.
Todeterminethequalityofthegage,astudyisperiodicallyperformedwherethe
gage(accept/rejectdecision)isappliedtoreferencepartswithknowndeviations
fromthedesiredspecifications.Thisprocessisdescribedintherespectivesection
oftheSTATISTICAElectronicManual,aswellastheAIAG(AutomotiveIndustry
ActionGroup)MeasurementSystemAnalysis(MSA)manual(2000).
Chapter2: StepbyStep Examples
Weareinterestedinevaluatingthegageperformanceforaprocessortypeof
manufacturedpartthatshouldbeidentifiedasunacceptable(shouldberejected),
whenitsreallowerlimitdropsbelow0.01(expressedhereasadeviationfromthe
spec).Inthedatafile,theAcceptanceprobabilitiessummarizethenumberof
referencepartsmeasurements,fromatotalof20suchpartsandmeasurements
each,thatweredeclaredasunacceptable(i.e.,thatwererejected).
Reviewing results.NowclickOKintheAttributegagestudy(Analyticmethods)
dialog.IntheResultsdialog,clicktheSummarybuttontoreviewthesummary
results.
Chapter2:StepbyStep Examples
Allimportantresultstodeterminethebiasandrepeatability(ofmeasurements)of
theattributegagearesummarizedonasinglepage.Fordetailsonthe
interpretationofthereportedstatisticsandgraphs,refertotheElectronicManual.
Example 6: STATISTICA Data Miner
STATISTICADataMiner(SDM)isacomprehensivesystemforpredictivemodeling
thatoffersawidevarietyofanalytictechniquesandmodelbuilding,validation,
andmodeldeploymentoptions.Thedefault,andperhapstheindustrystandard,
typeofuserinterfaceprovidedinSDMfollowsthegeneralinteractivedatamining
workspaceapproachthatenablesuserstobuildmodelsbydraggingicons
representingstepsofdataacquisition,datapreparation,modeling,and
deploymentandconnectthemwitharrows.Theworkspaceuserinterfaceoption
inSDMrepresentsapowerfulalternativetothetraditionalinteractivedata
analysisuserinterface,anditcanbeusednotonlyasatoolfordevelopingand
testingpredictivedataminingmodes,butalsoasapowerfulgeneraltooltobe
usedforvisualprogrammingofanalyticworkflowsformanytypesofanalyses.
Ablankdataminingworkspacewillbedisplayed.
Now,click onthetoolbartodisplaytheSelectDataSourcedialog,
usedtoselectadatafileforanalysis.Next,theSelectdependentvariablesand
predictorsdialogisdisplayed;clickthe buttontodisplaythevariable
selectiondialog,usedtospecifythedependentvariablesandpredictors.Then,
click tocreateanalyticnodes,andconnectthemwith arrows
tospecifythedesiredprojectworkflow.
ThefollowingsectionincludesastepbystepexampleofDataMinerRecipesan
innovativeuserinterfacefordataminingintroducedbyStatSoftwhichoffersa
powerfulalternativetotheworkspacebasedapproachtomodelbuilding,andcan
beusedbybothnovicesandadvancedanalysts.
Overview
ThisexamplepertainstoSTATISTICADataMinerRecipes,aStatSoftproductthat
offersawideselectionofmethodsforpredictivedatamining.
Chapter2:StepbyStep Examples
Inthisapproach,yousimplyfollowarecipelikeuserinterfacetocompletethe
necessarystepstomovetoasolution.Infact,mostofthesestepsareentirely
automatedsothattheonlyrequiredinputistodefinethedataandvariablesfor
theanalyses,whiletheprogramautomaticallydoestherestdetermineslearning
andtestingsamples,performsfeatureselection,triesvariousdatamining
algorithmsandmethods,andevaluatesresultstoselectthebestdatamining
model.Thesecomputationsandanalysescanbeperformedwitheitherthe
desktopSTATISTICADataMinersoftwareor,ifavailable,ontheSTATISTICAData
MinerServer.
Data Miner Recipes Project Files
WhenyousaveaDataMinerRecipesprojectatanystageofcompletion,two
separatefilesarecreated:
ADataMinerRecipesfilewiththefilenameextension.dmrproj
ASTATISTICAWorkbookfilebythesamename,butwiththefilename
extension.stw,containingresultsanddetailedinformationforeachstepof
therecipe
Chapter2: StepbyStep Examples
Theresultsstoredinthisworkbookprovidecompletedocumentationforthe
computationsandanalysesperformedastheDataMinerRecipewas(orisinthe
processofbeing)completed.Therefore,ifthedatamininganalysesareperformed
inaregulated(e.g.,FDA,ISO,etc.)environment,orifdataminingispartofan
organizationsmissioncriticalactivitiesperformedundertheguidanceandin
compliancewithspecificstandardoperatingprocedures(SOPs),thenitisusually
recommendedthatthisfilebestoredintheSTATISTICADocumentManagement
SystemalongwiththeDataMinerRecipeprojectfile(.dmrproj).
Using STATISTICA Data Miner
Recipes (SDMR)
Thisexampleillustrateshowquicklyandefficientlydataminingprojectscanbe
completedusingSTATISTICADataMinerRecipes,evenifthebestsolutiontothe
(prediction)problememergesonlyafter(automatically)comparingtheefficacyof
variousadvanceddataminingalgorithms.
Inthisexample,wewillexploretheuseofSDMRforcreditscoringapplications.
TheexampleisbasedonthedatafileCreditScoring.sta,whichcontains
observationson18variablesfor1,000pastapplicantsforcredit.Eachapplicant
wasratedasgoodcredit(700cases)orbadcredit(300cases).Wewantto
developacreditscoringmodelthatcanbeusedtodetermineifanewapplicantis
agoodcreditriskorabadcreditrisk,basedonthevaluesofoneormoreofthe
Chapter2: StepbyStep Examples
ThestepnodepanelislocatedintheupperleftareaoftheStepstab.Itcontains
fourmajornodes:Datapreparation,Dataforanalysis,Dataredundancy,and
Targetvariable.
Nodes (steps).Eachnode(orstep)canexistinoneoffourstates,dependingon
whetherallrequiredoptionshavebeenspecified.Eachstateisrepresentedbyan
icon:ared
indicatesawaitstate,meaningastepcannotbestartedbecauseitis
dependentonapreviousstepthathasnotbeencompleted;ayellow indicatesa
readystate,meaningyouarereadytostartthestepbecausepreviousstepshave
beencompleted;agreen indicatesacompletedstep.Notethatyoumustclick
theNextstepbuttontochangetheyellow (readystate)tothegreen
(completedstate).Thechangewillbemadeonlyifthestephasbeensuccessfully
completed(i.e.,allrequiredinformationhasbeenspecified).Lastly,ifyouhave
openedadatasetandselectedvariables,andyoudonotwanttoproceedstepby
stepthroughalltheoptions,youcanselecttheConfigureallstepscheckboxon
theStepstab.Thestepswillnowberepresentedbyanavy icon.Youcanselect
anyofthestepsandmodifytheoptions,oryoucanleavealloptionsattheir
Chapter2:StepbyStep Examples
Then,clicktheOKbutton.
Chapter2: StepbyStep Examples
changestoagreen ).
Data for Analysis
AftertheDatapreparationstepiscompleted,theDataforanalysisstepwillbe
selectedautomatically.OntheDataforanalysistab,clicktheSelecttesting
samplebutton,andintheTestingSampleSpecificationsdialog,selectthe
Variableoptionbutton.Verifythatthecategory(value)Trainisenteredinthe
CodefortrainingsamplefieldandTestisenteredintheCodefortestingsample
field.
Then,clicktheOKbutton.Themodelswillbefittedusingthetrainingsampleand
evaluatedusingtheobservationsinthetestingsample.Byusingobservationsthat
didnotparticipateinthemodelfittingcomputations,thegoodnessoffitstatistics
computedfor(predictedvaluesderivedfrom)thedifferentdataminingmodels
(algorithms)canbeusedtoevaluatethepredictivevalidityofeachmodeland,
hence,canbeusedtocomparemodelsandtochooseoneormoreoverothers.
Descriptive statistics.Thisstepwillalsocomputedescriptivestatisticsforall
variablesselectedintheanalysis.Descriptivestatsprovideusefulinformation
aboutrangesanddistributionsofthedatausedfortheproject.
Chapter2:StepbyStep Examples
ClicktheOKbutton.Thedatacleaningandpreprocessingformodelbuildingisnow
complete.
Target Variable: Building Predictive Model
Next,weneedtobuildpredictivemodelsforthetargetinthisexample.Inthe
stepnodepanel,theTargetvariablenodehasabranchingstructurewiththe
parentnodeconnectingtofourchildnodesincludingImportantvariables,Model
building,Evaluation,andDeployment.
ClicktheOKbuttoninthisdialog,andthenclicktheNextstepbuttontocomplete
thisstep.Toreviewasummaryoftheanalysisthusfar,ontheStepstab,clickthe
Reportbutton,andfromthedropdownlist,selectSummaryreporttodisplaythe
Resultsworkbook.
Chapter2:StepbyStep Examples
Inthismatrix,youcanseethatthismodelpredicted68outof103badcredit
riskscorrectly,butmisclassified35ofthem.Thisinformationisusuallymuch
moreinformativethantheoverallmisclassificationrate,whichsimplytellsusthat
theoverallaccuracyis68.52%.
DisplaytheDataminerrecipesdialogagain,andclicktheNextstepbutton.A
messageisdisplayedwithinstructionstoselectonlyonemodelfordeployment.
ClickOK,andclearthecheckboxesadjacenttoC&RTandNeuralnetwork.Wewill
deploytheBoostingTreesmodelthatgaveusthebestpredictiveaccuracyonthe
testsample.Now,clicktheNextstepbuttonagain.
Deployment
ThefinalDeploymentstepinvolvesusingthebestmodelandapplyingittonew
datainordertopredictthegoodorbadcustomers.Thisstepalsoprovidesthe
optionforwritingbackthescoringinformation(classificationprobabilities
computedbythebestmodel,predictedclassification,etc.)totheoriginalinput
datafileordatabase.Thisisextremelyusefulfordeployingmodelsonverylarge
datasetstoscoredatabases.
OntheDeploymenttab,clicktheDatafilefordeploymentbuttonandopenthe
CreditScoring.stadatafile(locatedintheDatasetsfolderinstalledwith
STATISTICA).Fordemonstrationpurposes,weareusingthesamedatafilefor
deploymentofthebestmodel.
Chapter2:StepbyStep Examples
ClicktheNextstepbuttontoscorethisdatafileusingthebestmodel.Thescoredfile
withclassificationsandpredictionprobabilities(titledSummaryofDeployment)is
locatedintheDeploymentfolderintheprojectworkbookasshownbelow.
Summary
Thepurposeofthisexampleistodemonstratetheefficiencyofthedataminer
workflowimplementedinSTATISTICADataMinerRecipes.Withonlyafewclicks,
theprogramwilltakeyouthroughthecompleteanalyticprocessfromthe
definitionofinputdataandanalysisproblem,throughdatacleaningandpreparation
andmodelbuilding,allthewaytofinalmodelselectionanddeployment.
Chapter2: StepbyStep Examples
buttonintheVariablespecificationdialogto
displaytheFunctionBrowserdialog,whichcontainsthecompletelistofformulas
andoperators(=,+,>,and,or,etc.).
Example: Spreadsheet Formula
OpentheAdstudy.stadatafile.Wewillcreateanewvariablethatisthemeanof
variables3through25(i.e.,MEASURE01throughMEASURE23).
Doubleclickonthefirstblankvariableheader(aftervariable25).TheAddCases
and/orVariablesdialogwillbedisplayed.ClicktheOKbuttontoacceptthe
default,whichistoaddonevariable.
TheVariablespecificationdialogforthenewvariablewillbedisplayed.Inthe
Displayformatgroup,selectNumber.IntheLongnamefieldatthebottomofthe
dialog,enter:=mean(v3: v25).
ClicktheOKbutton.Adialogwillbedisplayedthatinformsyouwhetherthe
formulaisformallycorrect.ClicktheYesbuttontoproceed.Thenewvariableis
nowfilledwiththemeanofvariables3through25foreachcase.
Sinceyoucanrefertovariablesbytheirnamesortheirnumbers,theformulawe
justcreatedcouldalsobeexpressedas:=mean(MEASURE01:MEASURE23).
Chapter2:StepbyStep Examples
Theonlydifferencesinsyntaxbetweenthebatchtransformationformulasandthe
spreadsheetformulasisthesupportformultipleformulasinthebatchoption,and
thefactthatbecausethebatchformulasarenotattachedtoanyspecificvariable
(infacttheycanbefreelycopiedfromdatafiletodatafile),theycannotstartwith
anequalsign,butmusthaveatargetvariable(e.g.,v1=...orMeasure03=...)sothat
STATISTICAknowstowhichvariableeachformulashouldapply.Thereisalsoan
optiontodistributeallbatchformulasintotherespectivevariablesinthe
spreadsheetandsavethemwiththedatafile,effectivelyreplacingthe
spreadsheetformulas(ifthereareany).
FollowingarethecalculationsusedtocalculateBMIandtoconvertHeight(in)to
centimeters,andtheformulastoenterintheBatchTransformationdialog:
Chapter2: StepbyStep Examples
BMI=('weight(lb)'/'Height(in)'**2)*703
'Height(cm)'='height(in)'*2.54
IntheFormulasfield,enterthelistoftransformationformulastobeappliedtothe
activedataspreadsheet.Separateeachtransformationformulabyareturn(press
ENTERonyourkeyboard).
ClicktheOKbuttonintheBatchTransformationFormulasdialog.TheAddNew
Variables?dialogwillbedisplayed;clicktheYesbuttontoaddthetwonew
variablestotheCharacteristics.stadatafile.Amessagewillbedisplayedtoinform
youwhethertheexpressionsyouenteredintheBatchTransformationdialogare
correct.IftheyareOK,clickYestoproceed.STATISTICAcalculatestheformulas
andaddsthetwovariables,BMIandHeight(cm),tothespreadsheet.
Chapter2:StepbyStep Examples
FromtheStatisticsmenu,selectBasicStatistics/Tables.TheSelectExcelRange
fortheAnalysisdialogwillbedisplayed.
ThisdialogisdisplayedwheneveryouselectacommandfromtheStatistics,Data
Mining,orGraphsmenuafteropeninganExcelworksheetintheSTATISTICA
application.NotethatSTATISTICAhasdeterminedthelogicalspecifications,but
theseoptionscanbechangedifnecessary.Whenvariablenamesarenotincluded
withtheExcelworksheet,STATISTICAwillassignvariablenames:Var1,Var2,Var3,
etc.AswithSTATISTICAspreadsheets,allvaluesinacolumnwillbeusedforthe
selectedanalysisunlesscaseselectionconditionsarespecified.
Forthisexample,clicktheOKbuttonintheSelectExcelRangefortheAnalysis
dialogtoacceptthedefaults;thedialogwillclose,andtheReview/EditColumn
Typesdialogwillbedisplayed.
Chapter2:StepbyStep Examples
InSTATISTICA,youcandefinethedatatypeforthespecificcolumns.Datatypes
includenumeric,text,mixednumericandtext,andmissingdata.Emptycellsinan
Excelworksheetarealwaystreatedasmissingdata,andwhenanumericcolumn
containstextvalues,thosevaluesarealsotreatedasmissingdata.STATISTICA
providesdefaultdatatypesforallcolumnsbasedonthefirstfewrowsofdata(in
fact,youcancleartheReview/Modifycolumntypesbeforeimportingcheckbox
intheSelectExcelRangefortheAnalysisdialogbeforeclickingOKinthatdialog,
andtheReview/EditColumnTypesdialogwillnotbedisplayed).However,youcan
changethedefaulttypesifneeded:selectthenameofthecolumnyouwantto
changeandclicktheEditbutton(ordoubleclickonthenameofthecolumnyou
wanttochange)todisplaytheChangeImportColumnTypedialog,whereyoucan
specifythetypeyouprefer.
Forthisexamplewewillacceptthedefaults,soclicktheCancelbuttoninthe
ChangeImportColumnTypedialog,andclicktheOKbuttonintheReview/Edit
ColumnTypesdialog.AfteryouclickOK,theStartupPanelfortheselectedanalysis
orgraphwillbedisplayed(inthisexample,theBasicStatisticsandTablesStartup
Panel),andyoucanproceedwiththeanalysisasusual.
Example 3: Accessing Data Directly
from a SQL Server Database
STATISTICAprovidesaccesstovirtuallyalldatabases(includingmanylargesystem
databasessuchasOracle,Sybase,etc.)viaSTATISTICAQuery,accessiblefrom
Chapter2: StepbyStep Examples
Fromthisdialog,youcanchooseexistingdatabaseconnectionsordefinenew
ones.Forthisexample,wellcreateanewdatabaseconnection,soclicktheNew
buttontodisplaytheDataLinkPropertiesdialog.
Chapter2:StepbyStep Examples
YoucanchooseeithertheOLEDBproviderthatwassuppliedbyyourdatabase
vendor,oroneoftheMicrosoftdefaultOLEDBprovidersthatiscompatiblewith
yourdatabasesystem.
Forthisexample,wellusetheNorthwindsampledatabaseinstalledwith
MicrosoftSQLServer,soselectMicrosoftOLEDBProviderforSQLServerandclick
theNext>>button.TheDataLinkPropertiesdialogConnectiontabwillbe
displayed.
SelectaserverfromtheSelectorenteraservernamedropdownlist.
Then,selectthelogonoptionbuttonappropriatetoyourSQLServerNorthwind
databaseinstallation.SelecteithertheUseWindowsNTIntegratedsecurity
optionbutton,orselecttheUseaspecificusernameandpasswordoptionbutton
andenteraUsernameandPasswordintherespectivefields.
Next,selectNorthwindfromtheSelectthedatabaseontheserverdropdownlist.
Chapter2: StepbyStep Examples
Selectthisconnection,andclickOK.TheSTATISTICAQuerywindowwillbe
displayed,withallthedatabasetablesinthetreeviewontheleft.
RightclickontheOrderDetailstable,andfromtheshortcutmenu,selectAddto
addthetabletothetableviewpane(theupperrightpaneintheSTATISTICAQuery
window).Then,rightclickontheProductstable,andaddittothetableviewpane.
SincebothtablescontaintheProductIDfield,STATISTICAQueryautomaticallyjoins
thetwotablesonthiskey.
Chapter2:StepbyStep Examples
Toselectthefieldstoincludeinthequery,rightclickintheOrderDetailstablein
thetableviewpane,andfromtheshortcutmenu,selectSelectAllFields.Inthe
Producttable,selecttheProductNamefield.
ClickthePreviewDatatabinthelowerrightpanetodisplayapreviewofthe
query.
ClicktheSQLStatementtabtodisplaytheSQLStatementgeneratedbythequery.
ToreturnthedatatoaSTATISTICASpreadsheet,clickthegreenarrowonthe
STATISTICAQuerytoolbar.TheReturningExternalDatatoSpreadsheetdialogwill
bedisplayed,whereyoucancontrolwhetherthequerywillbeplacedintoanew
orcurrentspreadsheetandadjustotherqueryparameters.SelecttheNew
Spreadsheetoptionbutton,andclicktheRunNowbuttontorunthequery.Ifthe
Chapter2: StepbyStep Examples
NowthedatacanbeanalyzedwithanyoftheSTATISTICAtools.Notethatthe
spreadsheetretainsthedatabaseconnection,andyoucanrerunthequeryatany
time:selecttheDatatab,andintheManagegroupclickExternalData.Select
RefreshDatafromthedropdownmenu.YoucanalsopressF5onyourkeyboard
whenthespreadsheetisopen.
Example 4: Data Preparation
Cleaning and Filtering
Summary of Options for Data
Filtering/Recoding
Inpractice,mostofthetimerequiredtocompleteadataanalysisordatamining
projectisspentonthepreparationofdata.Sometimesasmuchas90%ofalltime
andeffortrequiredtocompleteaprojectisrelatedtothepropercleaningand
preparationofthedata.
Whenbuildingpredictionmodelsusingdataminingtools,orevenwhenjust
computingsimpledescriptivestatistics(averages,frequencydistributions),results
ofanalysescanbeverymisleadingif,forexample,largenumbersofduplicate
recordsareincluded(e.g.,thesamepartnumbersarerecordedmultipletimes)or
thedataincludeoutliersormiscodedvalues(outsidethevaliddataranges)or
excessivenumbersofmissing(blank)data.
OntheDatatab,intheTransformationsgroup,clickFilter/Recodetodisplaya
dropdownmenucontainingcommandstoaddresssuchdataqualityissuesquickly
andeffectivelysothatmeaningfulandvaliddataanalysesordataminingprojects
canbecompletedinlesstime.
Chapter2:StepbyStep Examples
IntheInputgroupbox,clicktheCasesbuttontodisplaytheSpreadsheetCase
SelectionConditionsdialog,whichcontainsoptionstoselectonlyspecified
observationsorcasesforthededupingoperations.Inthisexample,wewillfilter
allthecases,soclicktheCancelbuttonintheSpreadsheetCaseSelection
Conditionsdialog.
TheUsecasenamescheckboxisclearedbydefault;wewillleavethisoptionasis
forthisexample.Whenthischeckboxisselected,casenamesareusedasoneof
thebasesfordistinction,i.e.,STATISTICAwilltreatasduplicatesanycasesthat
havethesamecasename(providedthecasesmatchonanyotherspecified
Chapter2: StepbyStep Examples
Includingsuchsparselypopulated(withdata)variablesinananalysismayleadto
erroneousresults,orpreventyoufrombuildingpredictivemodelsaltogether
(dependingonhowthemissingdataarehandledlaterintheanalyses).Therefore,
youmaywanttoidentifysuchsparsevariablesaheadoftimeusingtheFilter
SparseDataoptions(accessiblefromtheFilter/RecodemenulocatedontheData
tabintheTransformationsgroup),andeliminatethemfromsubsequent
consideration.
Chapter2: StepbyStep Examples
Suchvariablesarenotusefulforpredictivemodeling,andtheProcessInvariant
Variablesoptions(accessiblefromtheFilter/RecodemenulocatedontheData
tabintheTransformationsgroup)enableyoutoidentifythosevariables
automatically,andexcludethemfromfurtheranalyses.
Recode Outliers
Extremedatavaluesoroutlierscangreatlyaffectvariousanalysesandcausepoor
accuracyofprediction(datamining)models.Thereisnoformaldefinitionofwhat
constitutesanoutlierorextremevalue,andSTATISTICAsgraphicaltoolsmay
providethebestwaytoreviewdatatoidentifysuchunusualobservations(e.g.,
youcouldcreateboxplotsofthekeyvariablestoidentifyextremeobservations
andbrushorflagtheminthedata).
Toautomaticallyprocesslistsofvariablestoidentifyandremoveoutliers,the
RecodeOutliersoptions(accessiblefromtheFilter/Recodemenulocatedonthe
Chapter2:StepbyStep Examples
Outlierscanberecodedtomissingdataortovaliddatavalues(e.g.,tothe
respectivepercentileboundaryvalues,etc.).
Process Missing Data
Missingdataorinvaliddatavaluesmustobviouslybedealtwithinamannerthatis
consistentwiththegoalsoftheanalyses.Insomecases,missingorinvaliddata
maythemselvesprovideusefulinformationaboutaprocessorvariableofinterest.
Forexample,inmarketingresearch,itiscommonthatrespondentswillrefuseto
providedetailedpersonalinformationregardingtheirhealth,financialstatus(e.g.,
savings),etc.,andsuchrefusalitselfmaybecorrelatedwithothersignificant
variablesofinterest(e.g.,refusaltoanswerquestionsrelatedtoincomemayitself
beagoodindicatorofhighincome,ifindeedwealthierindividualsinthesurvey
tendednottoanswerthosequestions).
Chapter2: StepbyStep Examples
TheProcessMissingDataoptions(accessiblefromtheFilter/Recodemenu
locatedontheDatatabintheTransformationsgroup)enableyoutorecode
missingdataflexibly,definemultiplemissingdatavaluesorcodesforasingle
variable(whichcanthenberecodedtothevariablemissingdatacode),orjustto
flagvariablesthathavemorethanacertainpercentageofmissingdata.
Imputation of Missing Data
(k-Nearest Neighbor)
Itisoftennotclearhowbesttorecodemissingdata,andinfact,sometimesby
recodingmissingdataforaparticularvariabletoaspecificvalue(e.g.,themean),
thefinalresultsmaybebiased.Forexample,supposeinasurveyallrespondents
whorefusetoreporttheirincometendtobeinthehigherincomebracket.Inthat
case,assigningthemeanincometothoseindividuals(i.e.,recodingmissingdata
forvariableIncometothemeanincomeforthewholesample)mayyieldhighly
misleadingresults.
STATISTICAincludesaveryefficientmethod(applicabletoverylargedatasetsand
databases)forreplacingmissingdatawithvaliddatavaluesthatareconsistent
withtheotherobservationsinthesample.Detailsregardingtheknearest
neighbormethodandalgorithmareprovidedintheElectronicHelpforthe
MachineLearningmoduleofSTATISTICADataMiner.
Inshort,usingtheMDImputationoptions(accessiblefromtheFilter/Recode
menu),inafirstpassthroughthedata,theknearestneighboralgorithmwillselect
Chapter2:StepbyStep Examples
Theknearestneighboralgorithmisfastandefficient,andprovidesaneffective
methodforreplacingmissingdataintheinputfilewithreasonableguesses
basedonsimilardatapointsinthesample.Thisapproachdoesnotmakeany
particularassumptionsaboutthenatureoftherelationshipsbetweenvariables
(i.e.,requirethatamodelbeestimatedforeachvariabletopredictmissingdata
values),butsimplyusestheobserveddataasthemodel.
Merge Data Files
TheSTATISTICAMergeOptionsdialogenablesyoutomergetwodatafileseither
bythevariablesorbythecasessothatyoucancentralizealloftheobservationsto
onetable.SelecttheDatatab,andintheManagegroup,clickMergetodisplay
theMergeOptionsdialog.
Chapter2: StepbyStep Examples
ClicktheHelp buttonintheupperrightcornerofthedialogtoaccessHelp
topicsdescribingalltheoptionsinthisdialog.
Creating Subsets
Ifyouhavealargespreadsheet,youmaywanttocreateanewspreadsheet
containingaspecifiedsubsetofthecurrentspreadsheet.Forexample,open
Boston2.sta.Thisdatasetcontainsoverathousandcases.Wewanttoextract
housingtractswithlowmedianprices.
SelecttheDatatab,andintheManagegroup,clickSubsettodisplaytheCreatea
Subsetdialog.
ClicktheCasesbuttontodisplaytheSpreadsheetCaseSelectionConditions
dialog,whichcontainsoptionstocreateconditionstodefinetheselectionofcases
tobeconsideredforthesample.
SelecttheEnableSelectionConditionscheckboxtoactivatetheoptions,andthen
selecttheSpecific,selectedbyoptionbuttonintheIncludecasesgroupboxto
specifywhichcasestoincludeintheanalysis.Typev1=LOWintheExpressiontext
box.
Chapter2:StepbyStep Examples
ClicktheOKbuttontosettheselectionconditionsandreturntotheCreatea
Subsetdialog,andclicktheOKbuttoninthisdialogtocreatethenewspreadsheet.
Theresultantspreadsheetcontains334cases(insteadoftheoriginal1,012cases)
andall15variablesfromtheoriginalspreadsheet.ForthePRICEvariable,allcases
haveavalueofLOW.
Example 5: Using STATISTICA ETL
(Extract, Transform, and Load)
TheSTATISTICAETL(Extract,Transform,andLoad)moduleprovidesunique
capabilitiesforprocessingandmergingdata,inparticular,processdatathatare
difficulttomanageusingstandarddatabasetools.ETLautomatestheprocessof
validatingandaligningmultiplediversedatasourcesintoasinglesourcesuitable
foradhocorautomatedanalyses.
ETLofferstwooptionsforaligningdata:Timeindexed,whichaggregatesdatafrom
multipledatasourcesbasedonadate/timestampvariableandalignsdataby
minute,hour,day,week,month,quarter,oryear;andIDbased,whichaggregates
datafrommultipledatasourcesbasedonanidentifiervariableandanoptional
timevariable,andoptionallyalignsdatabyNequalintervalsorNuserspecified
intervals.
ThisexampleillustrateshowtheETLmodulehandlesstockrelateddatasetswith
differenttimeintervals.Stocksareboughtandsoldatvaryingpricesthroughout
eachday.Microsoft(tickerMSFT)andOracle(tickerORCL)aresoftwarecompanies
Chapter2: StepbyStep Examples
ClicktheAdddatasourcebuttontodisplaytheSelectDataSourcesdialog.
Chapter2:StepbyStep Examples
ClicktheDocumentsbuttontodisplaytheSelectDocumentsdialog.Selectthe
OpenSpreadsheetsDocumentscheckboxtoselectbothdatafiles
(MicrosoftPrices.staandOraclePrices.sta).
ClicktheOKbuttonintheSelectDocumentsdialog,andthenclicktheOKbutton
intheSelectDataSourcesdialog.TheSTATISTICAExtract,Transform,andLoad
(ETL):TimeindexedStartupPanelwillappearasshownbelow:
SelectMicrosoftPrices.stainthefilelistatthetopofthedialog,andclickthe
VariablesbuttontodisplaytheSelectvariablesdialog.SelectDATEfromthe
Date/Timestamplist,andselectCLOSEfromtheVariableslist.
Foradditionaldate/timeoptions,selecttheOptionstab.SelecttheFilterallinput
datasourcesbythefollowingDate/Timecheckbox.Tolimitthedatathatis
returnedfrombothoftheselecteddatafiles,enter11/2/2007intheStartdate
fieldand12/28/2007intheEnddatefield.Thiswillreturneightweeksofdata
(FridaytoFriday).
Now,clicktheResultsbuttontomergethedataintoaspreadsheet.
Chapter2:StepbyStep Examples
Thetwodatafilesarenowalignedweeklybydatefortherange11/2/2007to
12/28/2007.ThedailyclosingMicrosoftpricesareaggregatedasmeans,whilethe
weeklyclosingOraclepricesareunchanged.
TheResultsspreadsheetdisplaysdate/timestampsascasesnamessothatthey
canbeusedforgraphingtheaggregatedandaligneddata.
SelecttheGraphstab.IntheMoregroup,click2DandselectLinePlots(Variables)
todisplaythe2DLinePlotsVariablesdialog.
ClicktheVariablesbutton,andinthevariableselectiondialog,selectvariables2
and3.Then,clicktheOKbutton.Inthe2DLineplotsVariablesdialog,select
MultiplefortheGraphtype,andclicktheOKbutton.Thefollowingimageshows
theresultantgraphplottingMicrosoftandOracleprices.
Afterspecifyingtheoptionsonthistab,clicktheOKbutton.
TheServertabhasnowbeenaddedtotheribbonbar.IntheUsergroup,clickLog
In,andenteryourusernameandpasswordifrequested.Uponsuccessfully
establishingaconnection,theoptionsontheServertabwillbecomeavailable.
TheOpen,Save,andSaveAscommandsintheFilegroupareusedtouploada
currentlyopenfiletotheserverordownloadafileandopenitlocally.Thereare
alsoexplicitoptionsintheTransfergrouptoDownloadFiletoandUploadFile
fromspecificfoldersontheserverandtheclient.
Note:Asrealworldexamplesoftimeorresourceconsuminganalysesareusually
basedonlargedatasetsand/orinvolveiterativealgorithmsrepresentedby
STATISTICAcomponentsthatarenotincludedinallconfigurationsofSTATISTICA,
wearedeliberatelygoingtouseanexamplethatdoesnotrequiremuchtimeto
complete.Buteveninasituationwhereasingleanalysisisquickandnotresource
intensive,youmightneedtorunafairlycomplicated,timeconsumingsequenceof
tasks,possiblyscheduledatcertaintimeintervals.Inthiscase,theSTATISTICA
EnterpriseServerschedulingfacilitiescouldbeusedonceyouhavecreatedand
uploadedacustomscriptthatrepresentstherequiredtasks(forexample,by
combiningthemacrosrecordedduringaSTATISTICAsession).
Now,recordasampleanalysismacro;forexample,completethestepsdescribed
inExample2:ANOVA(page34).
Aftercompletingtheexample,intheANOVAResultsdialog,clicktheOptions
button(locatedatthebottomofthedialog),andfromthedropdownlist,select
CreateMacro.IntheNewMacrodialog,acceptalldefaults,andclickOK.Testthe
Chapter2: StepbyStep Examples
Weneedtoselectatasktooffload(ascriptoraDataMinerproject)and,
optionally,adatasetonwhichthetaskwilloperate(thedatasetcouldbean
optionalcomponentsinceDataMinerprojectsmayhavetheirdatasetsembedded
andmacrosmightexplicitlyloaddatasetsornotrequirethematall).
Sincethereisanopenactivedataset(Adstudy.sta)andanopenSTATISTICAMacro
(oursampleanalysis),thedefaultsettingsoftheoptionsintheOffloadatask
dialogspecifytousethemforoffloading.Instead,thisexamplewilldemonstrate
howtoreferenceataskandaserversidedataset.Thisoptionisusefulsinceit
givesyoutheadvantageofcentralserversidestorage,whichisespecially
beneficialinthecaseoflargedatasets(possiblydynamicallyupdated)thatare
usedbymultipleusers.
Toreferenceaserversidedataset,intheDataSourcegroupbox,selecttheSelect
datafilestoredontheserveroptionbuttontodisplaytheSTATISTICAEnterprise
ServerRepositorydialog.
Chapter2:StepbyStep Examples
ThedirectorystructureinthetreeviewofthedialogrepresentstheSTATISTICA
EnterpriseServerRepository(possiblyabridgedaccordingtoyourparticular
permissions).ClickontheDatasetsfolderintheleftpane,andselectAdstudy.sta
intherightpane(oryoucanenterthepathintheeditboxatthebottomofthe
dialog).
ClickOKintheSTATISTICAEnterpriseRepositorydialogandintheOffloadatask
dialog.STATISTICAwillsubmitthetasktotheserver,uploadingfilesifneeded.
Nowyoucanswitchtootheractivities,whileperiodicallymonitoringthestatusof
offloadedtasksbyclickingStatusintheTasksgroupontheServertabtodisplay
theTaskStatusdialog.ThefollowingillustrationshowsaTaskStatusdialog
containingseveraloffloadedtasks.
ThetaskliststatuscanbeupdatedmanuallybyclickingtheRefreshbuttonor
automaticallybyselectingtheAutomaticcheckboxinthelowerrightportionof
theTaskStatusdialog.TasksgothroughPendingandRunningstatestoeither
CompletedorScriptError.
Chapter2: StepbyStep Examples
Notethatwhenspreadsheetaudittrailloggingisenabled,thespreadsheetis
automaticallysettodirectmode,i.e.,changesmadetothespreadsheetwillbe
immediatelywrittentodisk.Thus,whenaudittrailloggingisenabled,changesto
thedatafilecannotbeundone.
SelecttheRequireuserstoenterreasoncommentsforeachchangecheckboxto
requireuserstoexplaineachchangemadetothespreadsheet.
Chapter2: StepbyStep Examples
Thelogviewerdisplaysagridofinformationregardingtheauditedactions
includingthesequencenumber,timeofchange,thecomputerusedtomakethe
change,userinformation,thenatureofthechange,andthereasonforthechange.
Columnwidthsintheloggridcanbeincreasedanddecreasedusingstandard
Windowstechniques.TheSpreadsheetAuditTrailsaresavedandembeddedinto
eachrespectivespreadsheet.
Password encryption vs. locking.Aspreadsheetcanbepasswordencryptedso
thatitcannotbeopenedwithoutthecorrectpassword.Onlyuserswhoknowthe
Chapter2:StepbyStep Examples
EnterapasswordintheDocumentPasswordfield,andclicktheOKbutton.The
Passworddialogwillbedisplayed,whereyoureenterthepasswordtoconfirmit;
passwordsarecontextsensitive.
ClicktheOKbuttoninthePassworddialog,andclosethedatafile.Adialogis
displayedwhereyoucanchoosetosavethechanges;clicktheYesbuttonsothat
thepasswordwillbeencrypted.Thenexttimeanyoneattemptstoopenthis
spreadsheet,thePassworddialogwillbedisplayed,andthecorrectpassword
mustbeenteredbeforethespreadsheetwillopen.
Lock a Spreadsheet
Inordertomeetcompliancerequirements,itisnecessarytohavecontrolofthe
reliabilityofinputdata.Usingthespreadsheetlockingoptions,youcanprevent
changestoallspreadsheetfeatures,fromtheappearanceofthedata(i.e.,display
Chapter2: StepbyStep Examples
Here,youcanspecifywhichaspectsofthespreadsheetthatyouwanttolock.
Whenuserstrytochangealockedfeature,amessagewillbedisplayed,informing
themthatthespreadsheetislocked.
SelecttheSpreadsheetdatacheckboxtopreventchangestotheactualdata
containedinthespreadsheet.Userswillbeunabletochangethedatavaluesand
themissingdatacode.Theywillalsobeunabletoperformanydatamanagement
operationsthataffectthespreadsheet(e.g.,changethedatatypeorthelengthfor
textvariables).Ifthischeckboxiscleared,userswillbeabletoeditthedata(e.g.,
byupdatingqueriesandSpreadsheetFormulasorbysimplytypinginnewvalues).
SelecttheDisplayelements(fonts,formats,etc.)checkboxtoprohibitthe
modificationoffontsandformatsusedinthespreadsheet.Optionsforchanging
thefontsize,color,type,andstyle(i.e.,bold,underline)willbedimmed.
Additionally,theoptionsforapplyingspreadsheetlayouts(accessiblebyselecting
theFormattabandclickingLayoutsintheSpreadsheetgroup)willbeunavailable.
SelecttheCaseselectionandweightscheckboxtopreventusersfromchanging
caseselectionconditionsandcaseweightsforthelockedspreadsheet.Userswill
notbeabletotoggletheuseofselectionconditionsorchangethecurrently
Chapter2:StepbyStep Examples
Chapter2:StepbyStep Examples
Then,clicktheCommitChangesbutton locatedatthetopoftheapplicationon
theQuickAccesstoolbartosavethechanges.Amessagewillbedisplayedthat
informsyouthattheuserdoesnthavepermissiontologin.ClicktheYesbuttonto
continue.
Wewillnowcreateagroup,givethegrouppermissions,andassignthenewuser
tothatgrouptoallowtheusertohavepermissiontologontotheEnterprise
Manager.Withthismethod,anypermissionchangeswillonlyneedtobeapplied
tothegroupinsteadoftheindividualusers,makingmaintenanceofusersin
STATISTICAEnterpriseeasier.
2. Create a New Group
IntheUserAdministrationnode,selecttheGroupsfolder,andintheproperties
page,clicktheNewGroupbuttontodisplaytheoptionstocreateanewgroup.In
theNamefield,enterTestGroup1.IntheGroupMembersframe,selectthecheck
boxadjacenttoTestUser1.Thiswilladdthepreviouslycreatedusertothegroup.
IntheGroupPermissionsframe,selectthecheckboxesadjacenttoAnalysisAdmin
(AADM)andWebUser(WUSR).Inthetreeview,clicktheplussign adjacentto
the
TestGroup1nodetoexpandit,andselectAnalysismodules.Intheproperties
page,clicktheSelectAllbuttontoselectallofthemodulesintheAvailable
analysismodule(s)list.
Chapter2: StepbyStep Examples
ThiswillgiveusersofthisgrouppermissiontologontobothWebanddesktop
STATISTICAandrunalloftheavailableanalysesandreports.
ClicktheCommitChangesbutton
tosavethechanges.
Wehavenowcreatedthenecessaryuserandgroupsecuritytorunanalysesand
reports.Whencreatingthedata,analysis,andreportconfigurationsinthenext
steps,wewillassignthisgrouptothoseobjectstoallowonlyuserswithinthe
grouptorunthem.
3. Create a System View Node
NowwewillcreateaSystemViewnodetoholdthisexamplesdata,analyses,and
reportconfiguration.Inthetreeview,clicktheplussign adjacenttotheSystem
Viewnodetoexpandit.RightclickontheSTATISTICAEnterprisefolder,andfrom
theshortcutmenu,selectNewFolder.IntheFoldernametextboxinthe
propertiespage,enterTestExample1asthenewfoldersname.
Chapter2:StepbyStep Examples
ClickCommitChangestosavethechange.Thisfolderwillnowbeusedtohouse
thedata,analyses,andreportconfigurations.
4. Create a New Database Connection
RightclickontheDatabaseConnectionsnodeinthetreeview,andfromthe
shortcutmenu,selectNewDatabaseConnectiontodisplaytheDataLink
Propertiesdialog.
Forthisexample,wellusetheNorthwindsampledatabaseinstalledwith
MicrosoftSQLServer,soselectMicrosoftOLEDBProviderforSQLServer,andclick
Chapter2: StepbyStep Examples
ClicktheTestConnectionbuttontoattemptaconnectiontothespecifieddata
source.ApromptwillbedisplayedthatacknowledgesthattheTestconnection
succeeded.Ifitdidntsucceed,checkyouraccesspermissionstothefileand
ensurethatthesettingsarecorrect.Forexample,spellingerrorsandcase
sensitivitycancausefailedconnections.
ClickOKintheprompt,andclickOKintheDataLinkPropertiesdialog.Inthe
resultingpropertiespage,enterTestExampleConnection1intheNamefield.
Chapter2:StepbyStep Examples
Then,clicktheAccessPermissionsbutton.FromthelistofAvailableUsersand
Groups,selectTestGroup1,andthenclickthetoparrowbutton tomoveTest
Group1totheAccessPermissionslist.
Now,clicktheCommitChangesbutton.
WiththedatabaseconnectioncreatedtotheNorthwinddatabase,wewillnow
createadataconfigurationtoextractdatafromthedatabase.
5. Create a Data Configuration
RightclickontheTestExample1folderinthetreeview,andfromtheshortcut
menu,selectNewDataConfiguration.Inthepropertiespage,enterTestExample
1intheNamefield.ClickthearrownexttotheConnectionfield,andfromthe
dropdownlist,selectTestExampleConnection1.
Chapter2: StepbyStep Examples
ClicktheNextStepbuttontodisplaythenewqueryoptions.
Chapter2:StepbyStep Examples
DragtheOrderstablefromtheleftpaneintotheeditorviewer(theupperright
pane),andthenselect,inthefollowingorder,theOrderID,ShipVia,ShipCountry,
andFreightfields.
SelectthePreviewDatatabinthequerypropertiesview(lowerrightpane)and
clicktheRefresh
toolbarbutton(theredexclamationmark).Thiswilltestthe
querytoensurethatvaluesarebeingretrievedfromthedefinedquery.
ClicktheOrderIDrowtohighlightit,andthenclicktheEditbuttontodisplay
optionstoedittheOrderIDcolumn.ClicktheAutoUpdatearrow,andfromthe
dropdownlist,selectFirstupdatecolumn.Thisenablesyoutodetectchangesin
theOrderIDcolumn.Inaddition,thecolumnissorted.
ClicktheNextStepbuttontoedittheShipViacolumn.ClicktheFilteringbuttonto
displaythefilteringoptions,andselecttheEnabledcheckboxtoallowfilteringon
theShipViacolumn.
Chapter2:StepbyStep Examples
ClicktheNextStepbuttontoreturntoShipViacolumneditingoptions,andthen
clicktheNextStepbuttontoedittheShipCountrycolumn.ClicktheFiltering
buttontodisplaythefilteringoptions,andselecttheEnabledcheckboxtoallow
filteringontheShipCountrycolumn.ClicktheNextStepbuttontoreturntothe
ShipCountrycolumneditingoptions,andthenclicktheNextStepbuttontoeditthe
Freightcolumn.ClicktheTargetTypearrow,andfromthedropdownlist,select
VariableCharacteristic.Thisoptionwillmakethiscolumnavailabletoperform
packagedSPCanalyses(thisisthecolumncontainingthedatatobeanalyzed).
Next,clicktheNextStepbuttontodisplaytheAccessPermissionsoptionsforthis
object.FromthelistofAvailableUsersandGroups,selectTestGroup1,andthen
clickthetoparrowbutton tomoveTestGroup1totheAccessPermissionslist.
Nowthisdataconfigurationwillbeexecutable(butnoteditable)bytheusersof
TestGroup1.
Chapter2: StepbyStep Examples
ClicktheNextStepbuttontocontinuecreatingtheanalysisconfiguration(leaving
thedefaultnamethesameasthedataconfigurationforexpediencyonly).Click
theNextStepbuttononceagaintocontinueeditingtheanalysisconfiguration.
Chapter2:StepbyStep Examples
ThisoptionwillspecifythatSTATISTICApromptforfilteringonthosecolumnsthat
haveFilteroptionsinthedataconfiguration(if,whendefiningtheFilteroptions,
theyweresettoRequiredwhenfiltering,thisstepwouldnotberequiredasit
wouldalwaysforceafilteringpromptwhenrunninginthisexampleitwasnot
requiredtoforcefiltering).ClicktheCommitChangesbuttontosavethisanalysis
configurationtoSTATISTICAEnterprise.
7. Run the Analysis Configuration
ClosetheEnterpriseManager,andlogontoSTATISTICAastheTestUser1user
createdinStep1.SelecttheEnterprisetab,andintheEnterprisegroup,clickRun
Analysis/ReporttodisplaytheRunAnalysisorReportdialog(thisdialogmaybe
displayedautomaticallydependingonyourconfiguration).SelecttheTestExample
1analysis,andclicktheOKbutton;theSQLCriteriadialogwillbedisplayed.
ClicktheFinishbuttontocompletethefilteringstep,extractthedata,andperform
apackagedanalysisontheFreightcolumn.
toveryelaborateuserinterfacesofvirtuallyunlimitedflexibility:
Chapter2: StepbyStep Examples
SeeAppendixBSTATISTICAEnterpriseServer,page263,formoreinformation.
Bydefault,whenyouselectspecificoutputfromaresultsdialog,theoutput(a
spreadsheetoragraph)isdisplayedandthedialogisautomaticallyminimizedinto
itsrespectiveanalysisbuttonatthebottomofthescreen.Clickthatbutton(or
pressCTRL+R)todisplaythedialogagainandresumetheanalysis.
Aselectionofoptionspertainingtoanalysismanagementareavailableonthe
shortcutmenu(accessedbyrightclickingonananalysisbuttonontheanalysis
bar)relatedtoeachrespectiveanalysisbutton(asshownabove).
A useful hint for those with large screens. Ifyouhavealargescreen,youcan
turnoffthedefaultminimizationoftheanalysisdialogsandtakeadvantageofthe
factthatmostofthesedialogsaresmalland,thus,canremainontheworkspace
withoutinterferingwiththeviewingofanalysisresults.Youcanadjustthisoption
eitherforaparticularanalysis(cleartheAutoMinimizecommandontheanalysis
Chapter3:UserInterface
EachStartupPanelcontainsalistofthetypesofanalysesavailableinthat
particularmodule.Clickinganywhereoutsidethepanelautomaticallyminimizesit
asabuttonontheanalysisbar.Ifyoursystemincludesahighresolutionscreen,
youcanchangethisdefaultandkeeptheconsecutivedialogs(ineachanalysis
sequence)displayedontheworkspace.
Analysis specification and output selection (results) dialogs. Whenthe
desiredanalysisisselectedintheStartupPanel,theanalysisspecificationdialogis
displayed,inwhichyouselectthevariablestobeanalyzedandotheroptionsand
Chapter3:UserInterface
Insomesimpleanalyses(suchasDescriptiveStatistics,shownintheillustration
above),theanalysisspecificationdialogalsoservesasanoutputselectiondialog
whereyoucanspecifythetypeandformatoftheoutput(e.g.,specific
spreadsheetsorgraphs).Mostanalyses,however,haveaseparateanalysis
specificationdialogandresultsdialog.
Spreadsheet facilities for scenario (what-if) analyses and customized
appearance.STATISTICAprovidesyouwiththecapabilitytoappend
supplementaryinformationaboutvariablemeasurementtypesandcasestatesto
yourspreadsheets.Thismetadatacanbeusedtocreateamorecomprehensive
descriptionofyourdataset,facilitatewhatiftypesofexploratoryanalyses,and
customizetheappearanceofcasesingraphs.
Case states and brushing.Youcanassigncasestatestocasesinorderto
customizetheappearanceofpointsingraphicaldisplays,thusmakingitveryeasy
toidentifyinfluentialandinterestingpoints.Awideselectionofsymbolsand
colorsisavailabletocustomizetheappearanceofselectedpoints.Notonlycan
casestatesbeassignedinthespreadsheetbeforeagraphiscreated,theycanalso
beassignedinteractivelyinthegraphviatheBrushingfacilities(accessibleby
clickingtheBrushingbutton intheCustomizeGraphgroupontheEdittab
whenagraphisdisplayed).Thecasestatesassignedinthegraphpropagateback
tothespreadsheet.Theabilitytoassigncasestatesineitherthespreadsheetor
graphfurtherfacilitatestheexploratoryvisualanalysisofdata.
Chapter3:UserInterface
Copyright StatSoft, 2011
STATISTICAQuickReference133
Measurement types and automatic variable pre-screening. Themodelingor
measurementtypeofavariablecanbeexplicitlydefinedinordertoindicatewhat
analysesandgraphsareappropriateforsuchavariable.Thesemeasurementtypes
willmapdirectlytosubsequentanalysesandgraphs,identifyingappropriate
variablesineachcase(e.g.,variablesoftypecategoricalwillbepresentwithinthe
listofcategoricalpredictorsavailableinaFactorialANOVA).
Inallvariableselectiondialogs(suchastheoneshownabove),theShow
appropriatevariablesonlyoptionisprovided,whichenablesyoutoprescreenor
filtervariablesaccordingtotheirMeasurementType(specifiedintheVariable
specificationdialog,accessiblebydoubleclickingonavariableheaderina
spreadsheet);ifthattypeisAuto,thentheAutomaticvariableprescreeningand
classificationoptions(locatedintheAnalysis/GraphoptionspaneoftheOptions
dialog,accessiblebyselectingtheToolstabandclickingOptions)determinehow
STATISTICAwillautomaticallydeterminetheMeasurementType.
Auto filtering (cloaking variables and cases).Filtering(accessiblebyselecting
theDatatabandclickingAutoFilterintheTransformationsgroup)isaquickand
easywaytodisplayaspecificportionofthedatainyourspreadsheetwithout
sortingthedataorcreatingasubset.Whenavariableisfiltered,onlythevalues
thatmeetthespecifiedcriteriaaredisplayedinthespreadsheet.Casesthatdonot
meetthecriteriaarehiddenfromsightbutnotremovedfromthespreadsheet
(e.g.,inthespreadsheetshownbelow,onlythecasesforGENDER=MALEare
displayed).
Althoughhidden,theyarestillavailableforstatisticalandgraphicalanalyses.
Chapter3:UserInterface
button
intheanalysisorgraphspecificationdialogandselectOutputtodisplaythe
Analysis/GraphOutputManagerdialog.
Toaccessglobaloutputoptions,selecttheToolstab.ClickOptionstodisplaythe
Optionsdialog,andselectOutputManager.Or,selecttheHometabandclick
OptionsintheToolsgroup.Formoreinformation,seetheElectronicManual.
Features of Analyses
STATISTICAprovidesdirectaccesstoallstatisticalanalysesviatheStatisticstab:
andtheDataMiningtab:
andprovidesdirectaccesstoallgraphicalanalysisdialogsviatheGraphstab:
Chapter3:UserInterface
Copyright StatSoft, 2011
STATISTICAQuickReference135
Thesetabsareneverdisabled,i.e.,theyareavailablewheneveranyinputdata
documentisopen.
TheStatisticsandDataMiningtabsprovideaccesstoallavailableanalysistypes
withinSTATISTICA.TheGraphstabprovidesdirectaccesstoavarietyofcommonly
usedgraphtypes(e.g.,scatterplots,histograms,means/errorplots,etc.)aswellas
hierarchicalaccesstoallgraphtypesinSTATISTICAincluding2DGraphs,3D
SequentialandXYZGraphs,CategorizedGraphs,UserdefinedGraphs,BlockData
Graphs,InputDataGraphs,andMultiGraphLayouts.Comprehensivediscussions
ofallthevarioustypesofstatisticsandgraphsofferedbySTATISTICAareavailable
intheglossaryoftheElectronicManual.Seealso,AppendixC:STATISTICAFamily
ofProducts(page275)formoreinformationonallmembersofthecomprehensive
selectionofdataanalysisapplicationsfromtheSTATISTICAfamilyofproducts.
Using the analysis bar. TotakeadvantageofSTATISTICAsmultitasking
functionality(seeMultipleAnalysisSupport,page128),STATISTICAanalysesare
organizedasfunctionalunitsthatarerepresentedwithbuttonsontheanalysisbar
atthebottomoftheapplicationwindow(abovethestatusbar,seethenext
illustration,whereDescriptiveStatistics,ClusterAnalysis,andCanonicalAnalysis
arerunningsimultaneously).Consecutivebuttonsareaddedasyoustartnew
analyses.
Chapter3:UserInterface
(inthelowerleftcornerofthescreen)andselectDocuments.
IntheGeneraloptionspaneoftheOptionsdialog(accessiblebyselectingthe
ToolstabandclickingOptions),youcanspecifyhowmanyrecentlyused
documentstodisplay(thedefaultis16).Formoredetailedinformationabouteach
documenttype,seetheoverviewsforworkbooks,spreadsheets,reports,graphs,
andmacrosonpage169;forfurtherinformation,seetheElectronicManual.
Tabs related to types of active document windows.Eachofthemaintypesof
STATISTICAdocumentwindows(seepage137)managesdatainadifferentway
and,thus,offersdifferentcustomizationandmanagementoptions.These
differencesarereflectedinthetabsthataccompanyeachtypeofwindow.Menu
commandsandbuttonsforeachofthemaintypesofdocumentsaredescribedin
detailintheElectronicManual.
Thetabsthatareavailablewhenworkbooksareopendependonthetypeof
documentthatiscurrentlyselectedintheworkbook.Therefore,whenyouare
Chapter3:UserInterface
Copyright StatSoft, 2011
STATISTICAQuickReference139
editingaspreadsheet,graph,report,ormacrowithinaworkbook,thetabs
relevantforthatdocumenttypeareavailable.Whenyouselectanemptynode
intheworkbooktreepane,bydefault,theWorkbooktabisdisplayed.
User-defined toolbars. Inadditiontothevarietyoftoolbarsprovidedonthe
STATISTICAclassicmenus(ontheribbonbar,clickthe iconintheupperleft
cornertodisplaytheclassicmenus),youcanalsocreateuserdefinedtoolbars.
ThesetoolbarscanincludeanycommandavailableinSTATISTICA,aswellasspecial
controls(i.e.,fontname,fontsize,graphstyles,etc.).Thetoolbarscanbegiven
anynameandcanbedesignatedtoopendependingontheactivedocumenttype.
Also,youcancustomizealltoolbars(includingexistingtoolbars)byadding
commandsandspecialcontrols.
Tocreateatoolbar(oreditanexistingone)usetheoptionsontheToolbarstabof
theCustomizedialog,accessiblebyselectingCustomizefromtheToolsmenu.
Customizingatoolbarisaseasyasdraggingcommandsfromthedialogtothe
toolbar,asshownintheillustrationbelow.
Shapesandlocationsoftoolbarscanbeeasilyadjusted(e.g.,alltoolbarscanbe
dockedorfreefloating).Alloftheseoptionsmakeitpossibleforyoutocreate
uniquetoolbarsthatprovideyouwithaveryspecializeduserinterface.The
ElectronicManualincludessimpletofollow,stepbystepinstructionsonhowto
Chapter3:UserInterface
dropdownmenulocatedintheupperleftcorneroftheribbonbar,seepage
23forfurtherdetailsonboththeglobalOutputManagerintheOptionsdialog
andtheAnalysis/GraphOutputManagerdialog).Thereareanumberofwaysto
outputtotheWeb,dependingontheversionofSTATISTICAyouhave.SharePoint
isaccessiblefromwithinSTATISTICA,andSDMSisanadditionalproductavailable
fromStatSoft.
CHAPTER
4
4
Chapter4:OutputfromAnalyses
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference149
Forexample,selectionsofdocumentscanbeextracted(e.g.,dragcopiedordrag
moved)toareportwindowortotheapplicationworkspace(i.e.,theSTATISTICA
applicationbackgroundwheretheywillbedisplayedinstandalonewindows).
Entirebranchescanbeplacedintootherworkbooksinavarietyofwaysinorderto
buildspecificfolderorganization,etc.
Technicallyspeaking,workbooksareActiveXdocumentcontainers(seepage238
forinformationonActiveXtechnology,seealsotheElectronicManual).
Workbooksarecompatiblewithavarietyofforeignfileformats(e.g.,Office
documents)thatcanbeeasilyinsertedintoworkbooksandinplaceedited.
User notes and comments in workbooks.Workbooksofferpowerfuloptionsto
efficientlymanageevenextremelylargeamountsofoutput,andtheymaybethe
bestoutputhandlingsolutionforbothnovicesandadvancedusers.Itmight
appearthatonepossibledrawbackisthatusercomments(e.g.,notes)and
supplementaryinformationcannotbeastransparentlyinsertedintothestream
oftheworkbookoutputastheycanintraditional,wordprocessorstylereports,
suchasSTATISTICAReports(seethenextsection).However,notethat:
AllSTATISTICAdocumentscaneasilybeannotated,botha)directly,by
typingtextintographs,tables,andreports,andb)indirectly,byentering
notesintotheCommentsboxoftheDocumentPropertiesdialog(accessed
byselectingPropertiesfromtheStartbutton
dropdownmenulocated
intheupperleftcorneroftheribbonbar),and
Formatteddocumentswithnotesandcomments(intheformoftextfiles,
STATISTICAReportdocuments,WordPadorwordprocessordocuments,
etc.)caneasilybeinsertedanywhereinthehierarchicalorganizationof
outputinworkbooks.Moreover,suchsummarynotesorcomment
documentscanbemadenodesforgroupsofsubordinateobjectstowhich
thenoteisrelatedtofurtherenhancetheirorganization.
Saving workbooks as Web pages.Workbookscanbesavedas*.html(Web)files
byselectingSaveAsontheHometabintheFilegroupfromtheSavemenu,andin
theSaveAsdialog,choosingWebPage(*.htm;*.html)fromtheSaveastype
dropdownlist.SavingasaWebpagewillcreatean*.htmlfileinthespecified
directorythatcanbeopenedwithstandardinternetbrowserssuchasMicrosoft
Chapter4:OutputfromAnalyses
TheWebpageoutputcontainsan.htmlbasedtreecontrolthatenablesyouto
navigateanddisplaythevariousworkbookimages,similartotheactualworkbook.
2. STAND-ALONE WINDOWS
STATISTICAoutputdocumentscanalsobedirectedtoaqueueofstandalone
windows;theQueueLengthcanbecontrolledintheOutputManageroptions
paneoftheOptionsdialog(accessiblebyselectingtheToolstabandclicking
Options).
Thecleardisadvantageofthisoutputmodeisitstotallackoforganizationandits
naturaltendencytocluttertheapplicationworkspace(someprocedurescan
generatehundredsoftablesorgraphswithaclickofthebutton).
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference151
Oneoftheadvantagesofthiswayofhandlingoutputisthatyoucaneasilycustom
arrangetheseobjectswithintheSTATISTICAapplicationworkspace(e.g.,tocreate
multiple,easytoidentifyreferencedocumentstobecomparedtothenew
output).However,notethatinordertoachievethateffect,youdonotneedto
configuretheoutputaheadoftimeandgeneratealargenumberof(mostly
unwanted)separatewindowsthatcancluttertheworkspace.Instead,individual,
specificoutputobjectsdirectedtoandstoredintheothertwochannels
(workbooksandreports)caneasilybedraggedoutfromtheirrespectivetree
viewsontotheapplicationworkspaceasneeded.
3. REPORTS
Whenperformingananalysis,theultimategoalistocreatemeaningfuloutputin
ordertogainanunderstandingofthedata.Themannerinwhichtheoutputis
producedisimportantaswell.STATISTICAoffersavarietyofmethodstoproduce
reportsthataccommodatethediverseneedsofusers.
STATISTICA Reports
STATISTICAReports(formoreinformation,seepage180)offeramoretraditionalway
ofhandlingoutputwhereeachobject(e.g.,aSTATISTICASpreadsheetorGraph,oran
Excelspreadsheet)isdisplayedsequentiallyinawordprocessorstyledocument.
However,thetechnologybehindthissimpleeditoroffersyouveryrich
functionality.Forexample,liketheworkbook(seeSTATISTICAWorkbooks,page
148),theSTATISTICAReportisalsoanActiveXcontainer(forinformationon
Chapter4:OutputfromAnalyses
dropdown
menulocatedintheupperleftcorneroftheribbonbar;orbyselectingtheHome
tab,clickingOptionsintheToolsgroup,andselectingOutputManagerinthe
Optionsdialogtreeview).IntheMicrosoftWordOutputdropdownlist,select
eitherMultipleWorddocuments(oneforeachanalysis/graph),CommonWord
document(onesharedforallanalyses/graphs),or[SelectFile]tobrowsetoa
preexistingWorddocument.
AlthoughWorddocumentsdonotprovidethenavigationaltreeofaSTATISTICA
WorkbookorReport,theadvantagesinsendingoutputtoWorddocumentsare
many.BysendingresultstoaWorddocument,youhaveallthewordprocessing
featuresofWordavailable.Forexample,youcanattachtemplatestocreate
customizeddocuments,addtablesofcontentandindices,trackchanges,etc.
WheninsertingalargespreadsheetintoaWorddocument,STATISTICA
automaticallydetectshowmanyvariablescanfitoneachpageandpartitionsthe
spreadsheetintoseveralWordtables.Ifthespreadsheetusescasenames,those
nameswillbethefirstcolumnineachtable.
AdditionalbenefitsofsendingresultstoaWorddocumentincludeincreased
printingfunctionality(e.g.,printingtofiles,manualduplex)andtheabilitytosave
resultsasWebpages.
5. OUTPUT TO THE WEB
Knowledge Portal
STATISTICAEnterpriseServerReports,oranySTATISTICAReports(seeHTML
Reports,page154),canbedistributedthroughtheKnowledgePortal.The
Chapter4:OutputfromAnalyses
TocreateafolderinthePortaldirectorytocontainyourreports,selectthePortal
folder,andthenclicktheCreatebuttontodisplaytheExplorerUserPrompt
dialog.Intheeditfield,enterthenewdirectorynameofSamplePortalFolder,and
clickOK.Adialogwillbedisplayedconfirmingthatthedirectory/Portal/Sample
PortalFolderwascreated.ClicktheShowMyDirectorybutton,andyouwillbe
returnedtotheMyDirectorydialog.SelecttheShowEmptyDirectoriescheckbox,
andthenclicktheRefreshbutton.ExpandthePortaldirectorybyclickingthe+
nexttothatfolder,andthenewSamplePortalFolderwillbedisplayed.
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference157
Notethatyoucancontrolwhocanreadandwritetothisfolderbyselectingthe
SamplePortalFolder,clickingtheSecuritybutton,andusingtheoptionstosetthe
userandgrouppermissionsforthefolderappropriately.
Publishing Content from STATISTICA
Enterprise Server
Nowthatthefolderhasbeencreated,youcanaddanalysisresultstoitforPortal
userstoviewusingeitherSTATISTICAEnterpriseServerorSTATISTICA.
InSTATISTICAEnterpriseServer,startwithatypicalanalysis.FromtheFilemenu,
selectOpenDataSpreadsheet.IntheSelectDataSourcedialog,selectthe
Datasetsfolderintheleftpane,selectthedatafileAdstudy.staintherightpane,
andclickOK.
Chapter4:OutputfromAnalyses
IntheDescriptiveStatisticsspecificationsdialog,selectAllresultsintheDetailof
computedresultsreportedfield.
ClickOKtodisplaytheresultsforthisanalysis,consistingofseveralspreadsheets
andgraphs.
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference159
Now,topublishthispagesothatotheruserscanseeitfromtheKnowledgePortal,
clickthePublishbuttonintheupperrightportionofthewindow.ThePublish
Destinationdialogwillbedisplayed.HereyoucanselecttheSamplePortalFolder
thatyoucreated.Youalsocancontrolwhocanhaveaccesstothisparticularpage
byselectingtheIwanttodefinewhocanaccessthisoutputpagecheckbox.
ClicktheNextbutton,andthepagewillbesavedtotheselecteddestination.
Chapter4:OutputfromAnalyses
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference161
AfteryouclicktheOKbuttonintheOptionsdialog,notethatthereisanowa
ServertabdisplayedinSTATISTICAnexttotheHometab.Theonlycommandon
theServertabthatisavailableinitiallyisLogIn;selectthatcommand.Ifyouhave
enabledintegratedlogin(andyourWindowsaccountisenabledonthe
STATISTICAEnterpriseServer),youwillbeloggedinautomatically.Otherwise,you
willbepromptedforaSTATISTICAEnterpriseServerusernameandpassword.
Onceyouhaveloggedin,theothercommandsareavailableontheServertab.
Now,wewillcreateananalysisanduploadtheresultstotheKnowledgePortal.
OpentheAdstudy.stadatafile:selecttheHometab,clicktheOpenarrow,and
selectOpenExamplesfromthedropdownmenu;intheOpenaSTATISTICAData
Filedialog,doubleclickontheDatasetsfolder,andthendoubleclickonthe
Adstudy.stafiletoopenthatspreadsheetforuseinSTATISTICA.
Next,selecttheStatisticstab,andintheBasegroup,clickBasicStatisticsto
displaytheBasicStatisticsandTablesStartupPanel.SelectDescriptivestatistics.
ClickOKtodisplaytheDescriptiveStatisticsdialog.
Chapter4:OutputfromAnalyses
ThisisthedocumentwewanttopublishtotheKnowledgePortal.OntheServer
tabintheFilegroup,clickSaveAs.TheSTATISTICAEnterpriseRepositorydialog
willbedisplayed,containingalistoffoldersyoucanreferenceintheSTATISTICA
EnterpriseServer.OpenthePortalfolder,selectSamplePortalFolder,andclickthe
OKbutton.ThiswilluploadtheworkbooktothatKnowledgePortaldirectory.
YoucanreviewthedocumentfromwithinSTATISTICAbyopeningabrowser
windowinsideoftheSTATISTICAworkspace.OntheServertabintheToolsgroup,
Chapter4:OutputfromAnalyses
Copyright StatSoft, 2011
STATISTICAQuickReference163
selectOpeninBrowser,andanewbrowserwindowwillbeopened,allowingyou
tologontotheSTATISTICAEnterpriseServer.
FromtheSTATISTICAEnterpriseServerFilemenu,chooseMyDirectory
Operations;inMyDirectory,youcannavigatetotheSamplePortalDirectory,and
seetheWorkbook1.stwfilethatwasuploaded.SelectthisfileandclicktheView
button,andtheworkbookwillbeopenedwithinthebrowser.
6. SHAREPOINT OR STATISTICA
DOCUMENT MANAGEMENT SYSTEM
(SDMS)
WithSTATISTICA,youcanalsorouteoutputtoeitherMicrosoftSharePointorto
theSTATISTICADocumentManagementSystem(SDMS).
SharePoint
WithSTATISTICASharePointintegration,youcanopen,checkout,checkin,and
uploadnewSTATISTICAfilestoSharePoint.
ToopenadocumentinSTATISTICAthatislocatedinSharePoint,selecttheHome
tab.ClicktheOpenarrow,andselectOpenDocument.IntheOpendialog,inthe
Lookindropdownlist,selecttheWebFoldertotheSharePointserverlocation
Chapter4:OutputfromAnalyses
Beforeusingtheseoptions,youmustfirstcreateaWebFoldertotheSharePoint
serverlocation.Todothis,clicktheStartbuttoninthelowerleftcornerofthe
Windowstaskbar,andclickComputer.Rightclickinanyopenareaintheright
paneoftheComputerdialog,andfromtheshortcutmenu,selectAddanetwork
locationtodisplaytheAddNetworkLocationdialog.ClicktheNextbutton.
DoubleclickChooseacustomnetworklocation.IntheInternetornetworkaddress
field,entertheWebaddressofyourSharePointlocation:https://sharepoint...,or
clicktheBrowsebuttontobrowsetoandselectthelocation.ClickNext.
LogontoSharePoint,andclickOK.EnteranamefortheWebFolderintheTypea
nameforthisnetworklocationfield,andclickNext.YouwillseeCompletingthe
AddNetworkLocationWizard;selecttheOpenthisnetworklocationwhenIclick
Finishcheckbox,andthenclickFinish.ANetworkLocationWebFolderhasbeen
createdintheNetworkLocationsectionofComputerwiththelabelyouchose.
STATISTICA Document Management
System (SDMS)
STATISTICADocumentManagementSystem(SDMS)isacompletedatabase
solutionpackageformanagingdocuments.SDMSenablesyoutoquickly,
efficiently,andsecurelysavedocumentsofanytypetoasecurerepository
database,andthenmanagethem[e.g.,findthem,accessthem,searchfor
content,review,organize,edit(withtrailloggingandversioning),approve,etc.].
Chapter4:OutputfromAnalyses
TheintuitiveuserinterfaceofSDMSmakesiteasytoperformalldocument
managementoperationsfromanycomputeronyournetworkorevenviathe
Internet.
IntheSTATISTICADocumentManagementSystem,everythingisdocumentedand
traceable.Forexample,documentsareneverdeleted.Whenadocumentisedited,
anewversionofthatdocumentiscreated,properlyauthenticated,andannotated
withelectronicsignatures.Authorizeduserscanberequiredtoexplicitlycheckout
thedocumentsfromtherepositoryandcheckthenewversionsintotherepository
withnotesanddocumentationregardingthenatureandpurposeoftheedits.
SDMSisspecificallydesignedtoensurecompliancewithFDA21CFRPart11
regulationsandSarbanesOxleylegislation,aswellasISO9000,9001,14001
documentationrequirements.
STATISTICADocumentManagementSystemseamlesslyintegrateswithall
STATISTICAproducts,fromdesktopandnetworkversions,toenterprisewide
installationssuchasSTATISTICAEnterpriseServerbasedworldwideinstallationsor
STATISTICAEnterprise/QC(forprocessanalysisandqualitycontrol/improvement).
SDMScanalsobeusedasastandalonesystem.
SDMSishighlyconfigurable,anditsfunctionalityiscompatiblewithotherapplications,
sothesystemcanbecustomizedtoaccommodateyourspecifictasksandcanbe
integratedseamlesslyintoexistingsystemsfordataanddocumentmanagement.
STATISTICA
DOCUMENTS
Workbooks ............................................................................................. 169
Spreadsheets (Multimedia Tables) ...................................................... 173
Reports ................................................................................................... 180
Graphs .................................................................................................... 182
Macros (STATISTICA Visual Basic Programs) ..................................... 183
STATISTICA Projects .............................................................................. 184
CHAPTER
5
5
Chapter5:STATISTICADocuments
Technicallyspeaking,STATISTICAWorkbooksareoptimizedActiveX(seepage238)
containersthatcanefficientlyhandlelargenumbersofdocuments.The
documentscanbeorganizedintohierarchiesoffoldersordocumentnodes(by
default,oneiscreatedforeachnewanalysis)usingatreeview,inwhichindividual
documents,folders,orentirebranchesofthetreecanbeflexiblymanaged.
Forexample,selectionsofdocumentscanbeextracted(e.g.,dragcopiedordrag
moved)tothereportwindowortotheapplicationworkspace(i.e.,theSTATISTICA
applicationbackgroundwheretheyaredisplayedinstandalonewindows).
CHAPTER
5
5
Chapter5:STATISTICADocuments
Displayingtabscanalsobesuppressedtosavespace.UnlikemanyExplorerstyle
navigationandorganizationapplicationsthatonlyallowfolderstohavechildren,
theSTATISTICAWorkbookallowsanyiteminthetreetohavechildren.For
example,youcanaddaspreadsheettoyourworkbook,andthenaddallthe
graphsproducedusingthedatainthespreadsheetaschildrentothespreadsheet.
AvarietyofdraganddropfeaturesandClipboardproceduresareavailabletoaid
youinorganizingtheworkbooktree.
TheworkbookcanholdallnativeSTATISTICAdocumentsincludingspreadsheets,
graphs,reports,andmacros.ItcancontainothertypesofActiveXdocumentsas
well,includingExcelspreadsheets,Worddocuments,andothers.Ifyouwantto
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference171
editthesedocuments,youcandosousingtheworkbookviewerpane.Toedita
Worddocument,doubleclickontheobjectintheworkbooktree.TheWord
documentopensintheviewer,andtheworkbookmenubarmergeswiththe
Wordmenubargivingyouaccesstoalloftheeditingfeaturesyouneed.
Workbookscanalsobeusedtostorealloutputfromaparticularanalysis.
Navigating the Workbook Tree
Theworkbooktreedisplaystheorganizationoffilesandfoldersintheworkbook,
displayedinanExplorerstyleformat.Itemswithplussignsnexttothemindicate
foldersorfilesthathavechildrenassociatedwiththem.Toexpandthetreefora
particularfolderorfile,clicktheplussignnexttoit.Theworkbookcansupportan
unlimitednumberoflevels,andindividualitemsfromthetreevieworentire
branchescanbeflexibly(interactively)managed(e.g.,draggingtocopyormove
betweenworkbooksorreports,etc.,orviatheshortcutmenu,asshownbelowin
thesecondimage).
Toselectaworkbookitemforrevieworediting,simplylocatethefileinthe
workbooktreeandclickonitsassociatedicon.Thedocumentwillbedisplayedin
theworkbookviewerpane.Notethatyoucanalsonavigatethroughthechildren
ofthecurrentlyselectednodeusingthenavigationtabsavailable(bydefault)at
thebottomoftheworkbookviewer.Youcaneasilymovethesenavigationtabsto
thetop,right,orleftoftheworkbookviewerbyrightclickingononeofthetabs
andselectingadifferentlocationfromtheshortcutmenuorselectingthe
appropriatecommandfromtheWorkbooktab,Toolsgroup,TabControlmenu.
Chapter5:STATISTICADocuments
ThesecommandsarealsoaccessibleontheWorkbooktab.
Theworkbooktreecanbeorganizedandmodifiedusingdraganddropfeatures
(aswellasClipboardprocedures).Usekeysonyourkeyboardtospecifywhether
anitemistobemovedorcopied,andwhetheranitemistobeinsertedasachild
(i.e.,onelevelbelow)orasasibling(i.e.,onthesamelevel).
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference173
Thefollowingtableillustratesfourdraganddropoptions:
Action Key Press Cursor Effect
MoveChild (none)
Movethefirstselecteditemonelevelbelow
thesecondselecteditem.
MoveSibling SHIFT
Movethefirstselecteditemdirectlybelow
andonthesamelevelasthesecond
selecteditem.
CopyChild CTRL
Copythefirstselecteditemonelevelbelow
thesecondselecteditem.
CopySibling SHIFT+CTRL
Copythefirstselecteditemdirectlybelow
andonthesamelevelasthesecond
selecteditem.
First,selecttheitem(s)thatyouwanttomoveorcopy.Dragtheselectiontoits
newlocationanddropit.Toselectasingleitem,clickontheitem(e.g.,
spreadsheet,graph,orreport).Toselectaparentnodeandallofitschildren,click
onthefolder.Notethathorizontaland/orverticalscrollingwithintheworkbook
treecanbeutilizedduringadraganddropoperation.
SPREADSHEETS
(MULTIMEDIA TABLES)
STATISTICASpreadsheetsarebasedonStatSoftsproprietarymultimediatable
technologyandareusedtomanagebothinputdataandthenumericortext(and,
optionally,anyothertypeof)output.Thebasicformofthespreadsheetisasimple
twodimensionaltablethatcanhandleapracticallyunlimitednumberofcases
(rows)andvariables(columns),andeachcellcancontainavirtuallyunlimited
numberofcharacters.Sound,video,graphs,animations,reportswithembedded
objects,oranyActiveXcompatibledocumentscanalsobeattached.
Chapter5:STATISTICADocuments
BecauseSTATISTICASpreadsheetscanalsocontainmacrosandanyuserdefined
userinterface,thesemultimediatablescanbeusedasaframeworkforcustom
applications(e.g.,withalistboxofoptionsoraseriesofbuttonsplacedinthe
upperleftcorner),selfrunningpresentations,animations,simulations,etc.
Title bar.Thetitlebardisplaysthenameofthespreadsheetfollowedbythe
spreadsheetextension(.sta).Ifthespreadsheetisaninputspreadsheet,thetitle
baralsodisplaysthenumberofvariablesbynumberofcases(e.g.,25vby50c).In
theimageshownabove,thetitlebarcontainsthetextData:Adstudy.sta(25vby
50c).
Info box.Youcanselecttheentirespreadsheetbyclickingonceinthelowerright
corner(themousepointerwillbethedefaultarrow)oftheinfobox,whichis
locatedintheupperleftcornerofthespreadsheetwindow.Toselecttheinfobox
only(forformatting),clickonceintheupperleftcorneroftheinfobox(themouse
pointerwillbeanoutlinedplussign ).Doubleclickintheinfoboxtoenteroredit
thetextintheinfobox(e.g.,additionaldetailsaboutthespreadsheet).Inthe
imageshownabove,theinfoboxcontainsthetextResponses(Peoria,IL).
Header.Theheaderislocatedimmediatelyabovethevariableheadersatthetop
ofthewindow.Doubleclicktheheadertoenteroredittextinformation.Toselect
theheaderonly(forformatting),clickonceintheupperleftcorner(themouse
pointerwillbeanoutlinedplussign ).PressCTRL+ENTERorALT+ENTERtoenteranew
line(notethatyouneedtoextendtheheightofthefieldtoseenewlinesthatyou
areadding).Intheimageshownabove,theheadercontainsthetextAdvertising
EffectivenessStudy.
Case headers.Thesecells,locatedatthefarleftofthewindow,containheader
informationforeachcase.Doubleclickonanycaseheadercelltoenteroredit
textinformation.Toselectthecaseheaderonly(forformatting),clickonceonthe
leftsideofthecaseheader(themousepointerwillbeanoutlinedplussign ).To
Chapter5:STATISTICADocuments
Itisalsopossibletoleaveastandalonespreadsheetopenbutdesignateitas
unavailableforanalysis.Todothis,selectthespreadsheet,andcleartheInput
checkboxontheDatatabintheModegroup.NowSTATISTICAautomatically
defaultstothemostrecentlyselectedinputspreadsheetforanalysis,ignoringall
spreadsheetsthatarenotdesignatedasinputspreadsheets.
STATISTICA Spreadsheet
OLE DB Provider
InadditiontousingspreadsheetsasdatasourcesforanalysesinSTATISTICA,
spreadsheetscanalsosupplydatatootherdatabaseawareapplicationsbyusing
theStatSoftOLEDBProviderforSTATISTICASpreadsheets.ThisOLEDBdriveris
installedwithSTATISTICA,andallowsreadonlyaccesstodatainSTATISTICA
SpreadsheetsusingtheindustrystandardStructuredQueryLanguage(SQL).You
canaccesstheOLEDBProvideratanypointthesystemallowsyoutochoosea
databaseconnection,usingthestandardMicrosoftDataLinkProperties.
Toaccessthisfunctionality,selecttheDatatab.IntheManagegroup,click
ExternalDataandfromthedropdownlist,selectCreateQuery.IntheDatabase
Connectiondialog,clicktheNewbuttontodisplaytheDataLinkPropertiesdialog,
whereyouselectStatSoftOLEDBProviderforSTATISTICASpreadsheets.
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference179
ClicktheNextbuttontodisplaytheConnectiontab.
TheDataSourcefieldspecifiesthedirectorypathwherethespreadsheetis
located.Whencreatingthequery,youcanchooseindividualspreadsheetfiles
withinthatdirectory.ThefollowingexampleusesSTATISTICAQuery,andhas
definedaconnectiontotheSpreadsheetOLEDB,specifyingthepathofthe
STATISTICAExamplesfolder.Eachspreadsheetwithinthefoldershowsupasa
potentialtable.
Chapter5:STATISTICADocuments
UsingtheStatSoftOLEDBProviderforSTATISTICASpreadsheetsenablesyouto
provideSTATISTICASpreadsheetdatatoanyapplication(includingSTATISTICA
itself)thatcanusetheindustrystandardOLEDBinterfaceforqueryingdata.
REPORTS
Reports(brieflyintroducedonpage150)inSTATISTICAofferamoretraditional
wayofhandlingoutput(comparedtoworkbooks)aseachobject(e.g.,a
STATISTICASpreadsheetorGraph,oranExcelspreadsheet)isdisplayed
sequentiallyinawordprocessorstyledocument.
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference181
However,thetechnologybehindthissimplereportoffersyourichfunctionality.
Forexample,liketheworkbook,eachSTATISTICAReportisalsoanActiveX(see
page238)containerwhereeachofitsobjects(notonlySTATISTICASpreadsheets
andGraphs,butalsoanyotherActiveXcompatibledocuments,e.g.,Word
documents)isactive,customizable,andinplaceeditable.Reportsarestoredinthe
STRfileformat,whichisaStatSoftextensionoftheMicrosoftRTF(RichText
Format,*.rtf)format.STRfilessharetheRTFformattinginformationand
additionallytheyincludethetreeviewinformation(whichcannotbestoredinthe
standardRTFfiles).Hence,reportfilesarebydefaultsavedwiththefilename
extension*.str,buttheycanalsobesavedasstandardRTFfiles(inwhichcasethe
treeinformationwillnotbepreserved).
Theobviousadvantagesofthiswayofhandlingoutput(moretraditionalthanthe
workbook)aretheabilitytoinsertnotesandcommentsinbetweentheobjects
aswellasitssupportforthemoretraditionalwayofquicklyscrollingthroughand
reviewingtheoutputtowhichsomeusersmaybeaccustomed.Also,onlythe
reportoutputincludesandpreservesarecordofthesupplementaryinformation,
whichcontainsadetailedlogoftheoptionsspecifiedfortheanalyses(e.g.,
selectedvariablesandtheirlabels,longnames,etc.,dependingonthelevelof
supplementaryinformationspecifiedintheOutputManager,seepage25).
Theobviousdrawback,however,ofthesetraditionalreportsistheinherentflat
structureimposedbytheirwordprocessorstyleformat,thoughthatiswhatsome
usersofcertainapplicationsmayfavor.
Navigating the Report Tree
Thereporttreedisplaystheorganizationoffilesinthereport.Thefilesare
displayedinanExplorerstyleformat;however,unlikeworkbooksthatcansupport
anynumberoflevels,thereportsupportsonlyoneleveloffiles.
YoucanembedanytypeofSTATISTICAdocumentinareport,including
spreadsheets,graphs,andanalyses.InadditiontoSTATISTICAdocumenttypes,
youcanembedothertypesofActiveX/OLEobjectsinareport,includingExcel
spreadsheets,Worddocuments,bitmapimages,andothers.Toeditoneofthese
typesofembeddeddocuments,doubleclickonthedocument.Thefileopensin
Chapter5:STATISTICADocuments
Commandsforinserting,extracting,renaming,andremovingitemsfromthe
reporttreeareavailablefromthereporttreeshortcutmenu(accessedbyright
clickinganywhereinthetree,asshownabove).
GRAPHS
GraphsrepresentanotherdistinctivetypeofSTATISTICAdocuments,andthey
offerrichfunctionalitybothintermsofthevarietyofwaysinwhichgraphscanbe
createdinSTATISTICAandintheselectionofgraphcustomizationtools.
SimilartotheotherSTATISTICAdocuments,graphsareActiveXcontainers(see
page238),whichmeansthattheycancontainavarietyofcompatibledocuments
(e.g.,Visiodrawings,Adobeillustrations,Excelspreadsheets,etc.).STATISTICA
GraphsarealsoActiveXobjectsand,therefore,canbelinkedtoorembeddedinto
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference183
othercompatibledocuments(e.g.,Worddocuments)wheretheycanbeinplace
editedbysimplydoubleclickingonthem.
GraphsarediscussedinmoredetailinChapter6Graphs.
MACROS (STATISTICA
VISUAL BASIC PROGRAMS)
TheindustrystandardSTATISTICAVisualBasic(SVB)language(integratedinto
STATISTICA)offersanother(alternative)userinterfacetothefunctionalityof
STATISTICA,anditoffersincomparablymorethanjustasupplementary
applicationprogramminglanguagethatcanbeusedtowritecustomextensions.
NotethatSTATISTICAVisualBasicisnotMicrosoftVisualBasic6.0.StatSoftowns
andmaintainsthecodeforSTATISTICAVisualBasic.SVBiscompatiblewith
MicrosoftsVB.NET,MicrosoftsVisualBasicforApplications(VBA),andalsowith
MicrosoftsVisualBasic6.0(VB6).SVBscriptinglanguageisuniqueintermsofits
flexibilityandcompatibility,anditisalsoverypowerful.ItprovidesaccesstoVisual
BasicforApplications(usedforscriptingMicrosoftOfficeproducts)andaccessto
the.NETFrameworkwithinthesamefile(seeChapter10Programming
STATISITCAfrom.NET,page247).OtherAPIscanalsobeaccessedandleverage
theflexibilityofSVBsuchas,forexample,YahoosStockQuoteAPIorGoogle
AnalyticsAPI.SVBoffersapowerful64bitsolutionforsystemintegration,
expansion,andcustomdevelopment.
STATISTICAVisualBasictakesfulladvantageoftheobjectmodelarchitectureof
STATISTICAandisusedtoaccessprogrammaticallyeveryaspectandvirtuallyevery
detailofthefunctionalityofSTATISTICA.Eventhemostcomplexanalysesand
graphscanberecordedintoVisualBasicmacrosandlaterberunrepeatedlyor
editedandusedasbuildingblocksofotherapplications.STATISTICAVisualBasic
addsanarsenalofmorethan14,000newfunctionstothestandard
comprehensivesyntaxofVisualBasic,thuscomprisingoneofthelargestand
richestdevelopmentenvironmentsavailable.
Chapter5:STATISTICADocuments
STATISTICAMacroscanbesavedinseveralformats,dependingonhowyouintend
tousethem(seetheSTATISTICAVisualBasicPrimerandtheElectronicManualfor
moreinformation).YoucanalsocopythemtotheClipboardandpastetheminto
otherprogramsordocuments.
STATISTICAVisualBasicisdiscussedinmoredetailinChapter8(page219).
STATISTICA PROJECTS
WhenperformingstatisticalanalysesandworkingwithSTATISTICAdocuments,
youwilloftenhavemanydifferentwindowsopen,andevendifferentanalysesin
differentstagesofprogress.STATISTICAprovidesameansforsavingyour
workspace,includinganyanalysesinprogress.YoucancloseSTATISTICAatany
pointduringananalysis,andwhenyoulaterreopentheproject,thepreviously
openedfilesandinprocessanalyseswillberestored.
TosaveaSTATISTICAProject,selecttheHometab,clicktheSavearrowinthe
Projectgroup,andselectSaveProjectAstodisplaytheSaveSTATISTICAProject
dialog.
Chapter5:STATISTICADocuments
Copyright StatSoft, 2011
STATISTICAQuickReference185
Inthisdialog,specifythepathandfilenameoftheSTATISTICAProjectfile(a
projectsextensionis.spf).Youcanalsospecifywhatitemstoincludeinthe
project.AllSTATISTICAdocumenttypescanbeselected(Spreadsheets,Graphs,
Workbooks,Macros,Reports,DataMinerprojects,InPlaceDatabaseprojects,
Analyses,andAnalysisresults).ForthoseSTATISTICAdocumentsthatarealready
storedondisk,youhavetheoptiontoeitherLinktotheexistingdocumentfile,or
tostoreacopyofthedocumentwithintheSTATISTICAProjectfile(Embedthe
documentintheproject).
InadditiontoSTATISTICAdocuments,projectfileswillalsosaveallinprogress
analyses.Theprojectfilewillstoretherecordedscriptsthatareautomatically
createdwheneveryanalysisisrun.Whentheprojectisreopened,thescriptsfor
theanalysesarererunagainsttheoriginaldataandtheanalysesdialogsaremade
visibleagaininexactlythestatetheywerewhentheprojectfilewassaved.
Projectfilesareaconvenientwaytosendinprogressanalysisstepsandresults
backandforthbetweenusersifyouelecttoembedthesaveddocumentsinthe
projectfile.Oneusercanrunanalysestoacertainpoint,andthensavetheproject
fileandpassittoanotheruser,whocanopentheprojectfileandcontinueexactly
wherethefirstuserstoppedtheanalyses.
Unlessyouconfigureitotherwise,STATISTICAwillautomaticallydisplayaprompt
askingifyouwanttosaveaprojectfilewhenquittingtheprogram,andwill
Chapter5:STATISTICADocuments
Locatedatthebottomofgraphs,youllfindtheinteractivegraphicscontrols(see
thenextillustrations),whichenableyoutoadjustthetransparencyoftheplot
areasandmarkers,andtoscrollandpaninordertointeractivelyscalethegraph.
Morecontrolsarelocatedin3Dgraphstoenableinteractiverotation.Clickthe
wrenchiconadjacenttothesliderstodisplaytheGraphOptionsdialog.
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference193
Left:2DGraph
Below:EnlargedimageofPanning(scaling),Scrolling,
andTransparencyControls
Left:Sectiontobe
scalediscircled
Right:Scaledviewof
leftgraphscircledarea
Left:Scatterplotwith
denseconcentrationof
datapoints
Right:Transparency
Controlrevealshidden
trends
Left:PlotAreaTransparencyControlcircled;making
plotareastransparentallowsportionsoftheplotto
overlapwhilestillbeingvisible
InteractiveScrolling
InteractivePanning
Chapter6:Graphs
Thenumbersinasimilarpiechart,however,canrepresentresultsofcalculations.
Forexample,theslicesofthepiecanrepresentrelativefrequenciesof
observationsthatbelongtocertaincategoriescalculatedbyoneofthehistogram
Left:3DGraph;RotationControlscircled
Below:EnlargedimageofRotationandTransparency
Controls
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference195
orfrequencycategorizationprocedures(e.g.,numbersofyearswhentheSales
werebelow$10million,between$10and$20million,andabove$20million).
Regardlessofthemethodthatwasusedtocreateagraph(i.e.,regardlessof
wherethenumbersrepresentedinthegraphwereobtainedorhowtheywere
calculated),allSTATISTICAGraphcustomizationandmultigraphicsmanagement
facilitiescanbeusedtochangetheappearanceofthegraphorintegrateitwith
othergraphsordocuments.
Chapter6:Graphs
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference197
andprecisedrawings:
butalsopresentationqualitydiagrams,posters,businesscharts,andother
displays:
Chapter6:Graphs
NotethatallthesegraphsarealsoavailableontheGraphstab,fromthe
STATISTICAStartmenu
onthestatusbar,orbyclickingtheGraphsGallery
buttononanygraphspecificationdialog.GraphsofInputDatadonotofferas
manyoptionsasthecorrespondingGraphsmenugraphs;however,theyare
quickertoselectbecauseunlikeGraphsmenugraphs:
GraphsofInputDatacanbecalleddirectlyfromthespreadsheetshortcut
menus,
GraphsofInputDatadonotrequireyoutoselectvariables(thevariable
selectionisdeterminedbythecurrentcursorpositionwithina
spreadsheet),and
GraphsofInputDatadonotrequireyoutoselectoptionsfromany
intermediatedialogs(defaultformatsoftherespectivegraphsare
produced).
GraphsofInputDataprocessdatadirectlyfromthecurrentinputdatafile,and
theytaketheircuesastowhichvariablestousefromthecurrentcursorposition
(inanytypeofspreadsheet).
Forexample,ifyourightclickasinglecorrelationinaresultsspreadsheetand
createaScatterplotbygraph,STATISTICAgeneratesa2Dscatterplotusingthe
originalrawvaluesofthetwovariablesrepresentedbythatcorrelation(seethe
IntroductoryExampleonpage11foramoredetailedexample).
AlthoughthemostconvenientwaytoselectGraphsofInputDataisviathe
spreadsheetshortcutmenu,youcanalsoselectthemfromtheGraphstaborthe
STATISTICAStartmenu
.Eithermethodwilldisplayasubmenufromwhichyou
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference201
canchooseoneofthestatisticalgraphsapplicabletothecurrentvariable(i.e.,to
thevariableindicatedbythecurrentcursorpositioninthespreadsheet).
Ifthespreadsheethasamatrixformatoraformatwhereacursorposition
indicatesnotonebuttwovariables(asintheillustrationshowingacorrelation
matrix,below),thenpredefinedbivariategraphsforthespecifiedpairofvariables
willbedirectlyavailablefromtheGraphsofInputDatasubmenus.
Otherwise,i.e.,whenthecurrentcursorpositionindicatesonlyonevariableasina
tableofdescriptivestatistics(asshowninthenextillustration),andifyouselect
anyofthebivariategraphsinthemenu,STATISTICAwillpromptyoutoselectthe
secondvariable.Forexample,ifyouselectScatterplotby,theSelectsecond
variabledialogwillbedisplayed,whereyouspecifybywhichvariableMeasure05
isgoingtobeplotted.
Chapter6:Graphs
Notethatthesegraphsareentirelyindependentfromtheconceptofinputdata.
Theyprocessvalues(numbers)fromwhateveriscurrentlyselectedintheblock
andignorethemeaningofthosenumbers(e.g.,thenumberscanberawdataor
valuesofcorrelationcoefficients).Thesegraphsofferaneffectivemeansof
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference203
visualizing,exploring,andefficientlysummarizingnumericoutputfromanalyses
displayedinresultsspreadsheets(e.g.,histogramsofMonteCarlooutputscoresin
theSEPATHmodule,oraboxplotofaggregatedmeansfromamultivariate
multipleclassificationtableintheANOVAmodule).
AlthoughthemostconvenientwaytoselectGraphsofBlockDataisviathe
shortcutmenuassociatedwiththeblockselectedinaspreadsheet,Graphsof
BlockDataarealsoavailablefromtheGraphstabortheSTATISTICAStartmenu
.WhencreatingGraphsofBlockData,youcanselectfromdefaultgraphs(e.g.,
Histogram:BlockColumnsorLinePlot:BlockRows),oryoucancreateyourown
customgraphsforeithertheselectedcellsintherowsorcolumns,orofallcellsin
theselectedrowsorcolumns(i.e.,goingbeyondthevaluesthatareselectedinthe
block).
Default graphs. Usingthedefaultgraphs(thefirstsixcommandsontheGraphsof
BlockDatasubmenu,shownintheillustrationabove),youcancreatespecified
graphswithasingleclick.Forspecificinformationoneachdefaultgraph,referto
theElectronicManual.
Custom graphs.SelectanyofthefourCustomGraphcommandstodisplaythe
SelectGraphdialog,whichprovidesavarietyofoptionsforcreatingcustomized
graph.
Forspecificinformationoncustomgraphs,refertotheElectronicManual.
Customizing graphs.AswithmostfeaturesofSTATISTICA,GraphsofBlockData
arefullycustomizable.SelectCustomizeListfromtheBlockDataGraphsmenuto
displaytheCustomizeGraphMenudialog,whichprovidesoptionstoremove,
rename,oreditthecurrentlylistedgraphsaswellastoaddnew(userdefined)
graphstotheGraphsofBlockDatamenu.
Chapter6:Graphs
,andofferhundredsoftypesofgraphical
representationsandanalyticsummariesofdata.
Notethat,unlikeGraphsofBlockData(whicharealsoincludedonthistabin
ordertoofferafullcomplementofallgraphicaloptionsaccessiblefromasingle
control),allothergraphtypesfromtheGraphstabarenotlimitedtothevaluesin
thecurrentoutputspreadsheet.Instead,theyprocessdatadirectlyfromthe
currentinputspreadsheet,inthesamewaythe(previouslydiscussed)Graphsof
InputDatado.Theyrepresenteitherstandardmethodstographicallysummarize
rawdata(e.g.,variousscatterplots,histograms,orplotsofcentraltendenciessuch
asmedians)orstandardgraphicalanalytictechniques(e.g.,categorizednormal
probabilityplots,detrendedprobabilityplots,orplotsofconfidenceintervalsof
regressionlines).Whengeneratingthesegraphs,STATISTICAtakesintoaccount
thecurrentcaseselectionandweightingconditionsforthevariablesselectedtobe
plotted.
Graphsmenugraphsinclude2DGraphs,3DSequentialGraphs,3DXYZGraphs,
MatrixPlots,IconPlots,CategorizedGraphs,andUserDefinedGraphs.Notethat
theCommongroupontheGraphstabincludesthemostcommonlyusedtypesof
graphs(Histograms,Scatterplots,Mean/ErrorPlots,etc.),andtheMoregroup
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference205
containsacomprehensivelistofallgraphtypes.Seealso,TypesofGraphsMenu
GraphsintheElectronicManual.
GRAPH BRUSHING AND
CASE STATES
GraphsthatarecreatedfromtheGraphstabarehighlyinteractivewiththe
spreadsheetfromwhichtheywerecreated.Youcanidentifyandselectpointsin
thegraphandspecifythattheyaretobehighlightedinthesourcespreadsheet,
andviceversa.
Inadditiontoselectingpointsingraphsandspreadsheets,youcanidentify
propertiesofacaseinaspreadsheetthatwillbeusedwhenthegraphiscreated
fromthatdata.Thesepropertiesincludethepointmarkerstyleandcolor,and
whetherthepointistobeexcludedfromthegraphand/orfitcalculations.
Tostartbrushingwithinagraph,clickthebrushing
buttonontheEdittabintheCustomizeGraphgroup,or
rightclickinthebackgroundofagraphandselectShow
BrushingfromtheshortcutmenutodisplaytheBrushing
dialog,whichisshownintheillustrationtotheright.
WiththedefaultSelectionBrush,whichisSimple,youcan
drawarectangleonthegraphtoselectthepointscontained
intherectangle.Thefollowingillustrationdemonstratesthis
fortheexampledatasetAdstudy.sta,witha2Dscatterplot
ofMEASURE01byMEASURE02.
Notethattheupperleftthreepointshavebeenselectedby
thebrushingtool,whichhighlightsthepointsinthegraphas
wellasthecorrespondingcasesinthespreadsheetfrom
whichthegraphwascreated.
Chapter6:Graphs
Alternatively,insteadofusingtheBrushingfacilities,youcanselectcasesinthe
spreadsheet(clickonthefarleftsideofthecasename)andthecorresponding
pointswillbemarkedinthegraph,asshowninthefollowingillustration,where
thefirstfivecasesintheAdstudy.staspreadsheethavebeenselected.
Chapter6:Graphs
Copyright StatSoft, 2011
STATISTICAQuickReference207
Youcanspecifyspreadsheetcasestatesfromeitheraspreadsheetoragraph.Ina
STATISTICASpreadsheet,rightclickonacasenametodisplaytheshortcutmenu,
whichcontainscommandsincludingOff,Label,MarkedPoints,andCaseStates.
Similarcommandsareavailablefromtheshortcutmenudisplayedwhenyouright
clickonthepointsinagraph.Thegraphwillusetheseoptionswhendisplayingthe
pointsrepresentedbythiscase.Forexample,ifyouselectLabel,the
correspondingpointswillbelabeled,asshowninthenextillustration.Notethat
thespreadsheetcasesaremarkedwithacasestateicontoindicatethatthecase
pointsarelabeled.
Rightclickonacasename,andfromtheshortcutmenuselectCaseStatesEdit
CaseStatestochangethecasemarkerand/orcolor.
NotethattheselectionofpointsisavailableforgraphtypesotherthanScatterplots.
Forhistograms,brushing/selectingahistogrambarwillselectthecorresponding
pointstothatbarinthespreadsheet.Thesameistrueoftheboxesinboxplots.
Usingcasestatesandbrushingandselectingpointsisparticularlyusefulwiththe
HiddenandExcludedcasestatesoptions.First,tomaketheseoptionsavailable,
displaytheOptionsdialog(selecttheToolstabandclickOptions),andinthetree
viewselectNavigation/Defaults(locatedunderSpreadsheets).Clearthe
Chapter6:Graphs
Thespecializedgraphsaredescribedinthedocumentationfortheanalysesfrom
whichtheycanbeproduced;forinformation,refertotheElectronicManual.
CREATING GRAPHS VIA
STATISTICA VISUAL BASIC
STATISTICAgraphicaloptionscanalsobeaccessedprogrammaticallyusingthe
builtinSTATISTICAVisualBasic(SVB)orothercompatiblelanguages.Therefore,
therearenolimitstohowdeeplycustomizedyourSTATISTICAgraphscanbe,
becauseSVB(withallitspowerfulcustomdrawingtoolsaswellastheSTATISTICA
basedlibraryofgraphicsprocedures)canbeusedtoproducevirtuallyanygraphics
ormultimediaoutputsupportedbythecontemporarycomputerhardware.
AnapplicationwritteninSTATISTICAVisualBasiccanoperateongraphsinthreeways:
Createanewgraphandthenmodify,print,orsaveit;
Accessanexistinggraphandthenmodifyit;
Chapter6:Graphs
AswithallotherfunctionsinSTATISTICAVisualBasic,functionstoaccessthe
graphicslibraryofSTATISTICAcanbeeasilyincorporatedintoSTATISTICAVisual
BasicprogramsviaahierarchicallyorganizedFunctionBrowser.Itcontainsshort
descriptionsofallfunctionsandoptionsthatcanbeinserteddirectlyintothesource
codeofyourprogram(i.e.,intotheSTATISTICAVisualBasicEditor,seepage225).
FormoreinformationonaccessingthegraphicslibrariesofSTATISTICAviathe
STATISTICAVisualBasicprogramminglanguage,refertotheElectronicManual.
CUSTOMIZING
STATISTICA
Customization of the Interactive User Interface ................................ 213
Customization of Documents ............................................................... 214
Local vs. Permanent Customizations .................................................. 215
General Defaults .................................................................................... 215
Graph Customization ............................................................................. 217
Maintaining Different Configurations of STATISTICA ........................ 218
Customized Configurations for Individual Users on a Network ........ 218
CHAPTER
7
7
Alltheseandothergeneralsettingsareaccessibleregardlessofthetypeof
documentthatiscurrentlyactive(e.g.,aspreadsheetoragraph).Formore
informationaboutaspecificoptionspane,seetheElectronicManual(i.e.,pressF1
toviewtheSTATISTICAHelptopicdescribingtheoptionscurrentlydisplayed).
Switching between alternative sets of defaults (configurations).Optionsare
providedintheConfigurationsoptionspaneoftheOptionsdialogthatenableyou
tomaintainlibrariesofsettingsandswitchbetweenthemfordifferentprojects
(orusers).Forfurtherdetails,seeMaintainingDifferentConfigurationsof
STATISTICAonpage218andintheElectronicManual.
Chapter7:CustomizingSTATISTICA
Copyright StatSoft, 2011
STATISTICAQuickReference217
GRAPH CUSTOMIZATION
Interactive graph customization.ThecustomizationoptionsinSTATISTICA
graphicsincludehundredsoffeaturesandtoolsthatcanbeusedtoadjustevery
detailofthedisplayandassociateddataprocessing.Theseoptionsarearrangedin
ahierarchicalmanner,sothoseusedmostoftenareaccessibledirectlyvia
shortcutsbydoubleclickingorrightclickingonaspecificelementofthegraph.
Permanent settings and automation options.Theinitial(default)settingsofall
graphfeaturescanbeeasilyadjustedsothateventhedefaultappearanceand
behaviorofSTATISTICAGraphswillmatchyourspecificneedsand/orwillrequire
verylittleinterventiononyourpart.VariousaspectsofSTATISTICAGraphscanbe
permanentlyadjustedbyusing:
1.theOptionsdialog(selecttheToolstabandclickOptions),
2.thecomprehensivesystemofgraphstyles,
3.userdefinedgraphs,and
4.STATISTICAVisualBasic.
ThesefacilitiesarebrieflyreviewedinChapter6Graphs(page190).Formore
information,pleaserefertotheElectronicManual.
TherearenolimitstohowdeeplycustomizedyourSTATISTICAcustomgraphs
canbe,becauseSTATISTICAVisualBasic(withallitspowerfulcustomdrawingtools
aswellastheSTATISTICAbasedlibraryofgraphicsprocedures)canbeusedto
producevirtuallyanygraphicsormultimediaoutputsupportedbycontemporary
computerhardware.Thosecustomdevelopeddisplaysormultimediaoutputcan
beassignedtoSTATISTICAtoolbars,menus,ordialogsandbecomeapermanent
partofyourSTATISTICAapplication.
Chapter7:CustomizingSTATISTICA
STATISTICA
VISUAL BASIC
Recording STATISTICA Visual Basic (SVB) Macros (Programs) ........ 224
Example: Recording an Analysis .......................................................... 230
ActiveX Objects and Documents (A Technical Note) ......................... 238
CHAPTER
8
8
UsingtheRlanguagerequiresthatyouhaveRinstalledoneitherthesame
computerrunningSTATISTICAoracomputeraccessiblefromtheSTATISTICA
EnterpriseServerinordertouseitsspecializedroutinesandcapabilitiesto:
AddnewRbasedmodules
LeverageSTATISTICAssuperiorgraphics,flexiblespreadsheets,and
convenientworkbookcontainersforvariousdocumenttypestohandle
outputfromR
IntegrateRintoSTATISTICAEnterprisetomakespecializedRfunctionality
availableasreusableanalysistemplatesforusersnotfamiliarwiththeR
language,inasecure,rolebasedenterpriseanalysissystem
AddRbasedanalyticnodestoSTATISTICADataMiner,thusleveragingallR
capabilitiesinsideSTATISTICAandDataMinerworkspaces
BuildscalableRserversusingSTATISTICAEnterpriseServertohandle
securityandloadbalancing,andtotakeadvantageofmultipleprocessor
serverstorunRfordemandingand/orvalidatedenterpriseapplications
SeetheElectronicManualformoreinformationonthesescriptinglanguages.
Chapter8:STATISTICAVisualBasic
Chapter8:STATISTICAVisualBasic
Alsoavailableisaninteractivedialogeditorthatenablesyoutobuilddialogboxes.
Tosummarize,STATISTICAVisualBasicisnotonlyapowerfulprogramming
language,butitrepresentsaverypowerful,professionalprogramming
environmentfordevelopingsimplemacrosaswellascomplexcustom
applications.
Visual Basic from other applications.SVBprogramscanalsobedevelopedby
enhancingVisualBasicprogramscreatedinotherapplications(e.g.,Excel)by
callingSTATISTICAfunctionsandprocedures.
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2011
STATISTICAQuickReference227
YoucanthenselectanddragthespecificitemfromtheCommandslistontoany
menuortoolbar.Notethatasyourmousepointerhoversoveramenu,themenu
willexpand,enablingyoutoinserttheiteminanysubmenuaswell.Oncethe
macroisplacedonthemenuortoolbarwhiletheCustomizedialogisdisplayed,
youcanrightclickthemacroandchangetheappearanceandtextoftheitem,as
wellasaddicons.
Running Macros from a command line. WithSTATISTICA,youcanexecuteSVB
programsfromthecommandlinebyusingthe/RunMacro=commandline
parameter.Thesyntaxis:
statist.exe /RunMacro=macroname
wheremacronameisthefilenameofthemacro.Ifafullpathisnotspecified,
STATISTICAwillattempttorunthemacrofromtheapplicationscurrentlyselected
directory(whichisWindowsdefaultbehavior).
Ifthemacrodoesnotmaketheapplicationoranydocumentvisible(throughthe
Application.Visible = True,orsimilardocumentproperties),theSTATISITCA
instancewillautomaticallyshutdownwhencomplete.Iftheapplicationismade
visible,theapplicationwillremainvisibleafterthemacrocompletes,andyouwill
needtoshutdowntheprogram.
Chapter8:STATISTICAVisualBasic
ClicktheOKbuttontodisplaytheDescriptiveStatisticsdialog.
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2011
STATISTICAQuickReference231
ClicktheVariablesbuttontodisplaytheSelectthevariablesfortheanalysis
dialog.SelectvariablesMEASURE01throughMEASURE23byclickingMEASURE01
anddraggingtoMEASURE23,andthenclickOK.
IntheDescriptiveStatisticsdialog,selecttheAdvancedtab,andnotethe
numerousoptionsavailable.
Forthisexample,wewillleavealloptionsattheirdefault.ClicktheSummary
buttontodisplaythedescriptivestatisticsfortheselectedvariables.
Chapter8:STATISTICAVisualBasic
Torunthismacro,selecttheDebugtab,andintheRungroup,clickRun(orpress
F5onyourkeyboard).TheexactDescriptiveStatisticsresultsthatweregenerated
intheinitialanalysiswillbereproduced.
LookattheSVBmacroforamoment.Towardthetop,oneofthelinesis:
Set newanalysis = Analysis (scBasicStatistics, ActiveInputDataSet)
ThisistellingthemacrothatitisgoingtoruntheBasicStatisticsanalysis,andthat
itwillbeusingtheactivedataset,thatis,thespreadsheetthatiscurrently
selectedwhenthemacroruns.
Afewlinesfurtherdownisasectionthatstartswith:
Dim oAD2 As STABasicStatistics.BasDescriptiveStatistics
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2011
STATISTICAQuickReference233
andunderthatarepropertiessuchas:
.PairwiseDeletionOfMD = True
Thesepropertiescorrespondtoalltheoptionsthatwereavailableonthedifferent
tabsoftheDescriptiveStatisticsdialog.Everyoptioninthedialogisrepresented
byaproperty,andallthecurrentsettingsarerecorded.Ifyoudecidetoincludea
MedianandtheSumofeachofthevariables,itiseasytoaddthistotheSVB
macro;justfindthelinesthatread:
.Median = False
and
.Sum = False
andchangetheseto:
.Median = True
and
.Sum = True
Now,runthemacroagainbypressingF5.Anewresultsspreadsheetwillbeadded
totheworkbook,thistimewithnewcolumnsofMedianandSum:
Letskeepthemacrowindowopenandstartanewanalysisonthesamesample
dataset.SelecttheAdstudyspreadsheettobringittothefront.SelecttheGraphs
tab,andintheMoregroup,click2D.SelectNormalProbabilityPlotstodisplaythe
NormalProbabilityPlotsdialog.
Chapter8:STATISTICAVisualBasic
ClicktheVariablesbutton,andintheSelectVariablesforProbabilityPlotdialog,
selectvariablesMEASURE01throughMEASURE03.ClickOKtoclosethisdialog,
andclickOKintheNormalProbabilityPlotsdialog.ThreeProbabilityPlotgraphs
willbeplacedintheresultsworkbook,oneforeachofthethreevariablesthat
wereselected.
ThestepsoftheProbabilityPlotanalysiswererecordedjustastheywereforthe
DescriptiveStatisticsanalysis.Tocreateanewmacrowiththesesteps,bringthe
NormalProbabilityPlotdialogtothefrontbyclickingthatbuttonontheAnalysis
Barinthelowerleftofthescreen,clickthe button,andselectCreate
Macrofromthedropdownmenu.IntheNewMacrodialog,clickOK,andanew
SVBMacrowindowisopenedwiththerecordedProbabilityPlotscript.
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2011
STATISTICAQuickReference235
AswiththeDescriptiveStatisticsanalysis,alltheoptionsselectedinthe
ProbabilityPlotdialogarespecifiedaspropertieswithinthemacro.Forinstance,
tochangethisfromaNormalProbabilityPlottoaHalfNormalProbabilityPlot,
locatethefollowingline:
.GraphType = scProbNormal
andchangeitto:
.GraphType = scProbHalfNormal
Also,letsexpandthevariablestoincludevariableMEASURE04.Todothis,findthe
followingline:
.Variables = "3-5"
Thislinecorrespondstothevariablesselectedfortheplots.Sinceweselected
MEASURE01throughMEASURE03,andthesearevariablenumbers3through5
fromthedataset,thisstringwasrecorded.ToaddMESURE04(variablenumber6),
changethislineto:
.Variables = 3-6
NowrunthemacrobypressingF5.FournewgraphsareproducedasHalfNormal
ProbabilityPlotsforvariablesMEASURE01throughMEASURE04.
Thisexamplehasdemonstratedhowyoucanrunanyanalysis,andthencreatea
macrooftheanalysisthatcanbeeditedandrerun.Additionally,thisexamplehas
Chapter8:STATISTICAVisualBasic
Noticethatthereisaredarrowoneachworkbookfolder.Thisisanindicatorthat
thescriptthatproducedtheresultsinthatfolderhasbeenattachedtothefolder.
ThisenablesSTATISTICAtorerunorresumetheanalysis.
Torerunananalysis,rightclickononeofthefolderslabeledDescriptivestatistics
dialog,andfromtheshortcutmenu,selectRerunAnalysis.TheRerunAnalysis
dialogwillbedisplayed.
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2011
STATISTICAQuickReference237
HereyoucanchoosetoUseoriginaldatasourceorUsenewdatasource.The
latteroptiongivesyouthepowerfulabilitytocreatetemplatesthatcanthenbe
appliedtonewdatasources.Inadditiontospecifyingthedatasource,youcan
choosetoReplacecurrentfoldercontentsorOutputtonewfolder.Inthis
example,leavethedefaults,andclickOK.Youwillseethatthecontentsofthe
folderarebrieflydeletedandthenaddedagainastheanalysisisrerun.
Onepurposeforthisfeatureistheabilitytoupdate/rerunresultsfromcomplex
analysesifnewdataisenteredintothespreadsheet.Forinstance,ifthedatain
theopendatafileAdstudy.stahasbeenchangedandtheanalysisisrerun,thenew
resultswillbecalculatedwiththenewdata.
Theresumeanalysisfunctionalityenablesyoutobringananalysisbacktothe
pointbeforetheresultsweregenerated,allowingyoutoselectdifferentoptionsor
continueananalysisinprogress.RightclickthesameDescriptivestatisticsdialog
folder,andfromtheshortcutmenu,selectResumeAnalysis.TheResumeAnalysis
dialogwillbedisplayed.Thisdialogalsocontainsoptionstospecifytheinputdata
source(originalornew).TheOutputoptionsforthenewresultsaretoOutputto
currentfolder(asifthisisjustanextensionofthepreviousanalysis)orOutputto
newfolder(asifthisisabrandnewanalysis).
Leavethedefaultsastheyare,andclickOK.TheDescriptiveStatisticsdialogwill
bedisplayed,withalltheoptionssettowhatwasusedwhentheselectedoutput
wascreated.SincethedefaultwastoOutputtocurrentfolder,clickingthe
Summarybuttonwillgeneratenewoutputtothesamefolder.
Chapter8:STATISTICAVisualBasic
Copyright StatSoft, 2007
STATISTICA QuickReference 241
STATISTICA
QUERY
Overview ................................................................................................. 243
Quick, Step-by-Step Instructions .......................................................... 244
In-Place Processing of Data on Remote Servers
(The IDP Technology Option) .......................................................... 245
OLAP Cubes ............................................................................................ 246
Large Database Files ............................................................................ 246
CHAPTER
9
9
.
2.AfteryouhaveselectedadatabaseconnectionandclickedtheOKbuttonin
theDataLinkPropertiesdialog,youwillhaveaccesstoSTATISTICAQueryin
whichyoucancreateaSQLstatementbyspecifyingthedesiredtables,
fields,joins,criteria,etc.(viatheTable,Join,andCriteriamenus)tobe
includedinyourquery.
Chapter9:STATISTICAQuery
Copyright StatSoft, 2011
STATISTICAQuickReference245
3.Onceyouhavespecifiedaquery,selectReturnDatatoSTATISTICAfromthe
Filemenu.TheReturningExternalDatatoSpreadsheetdialogwillbe
displayed,inwhichyoucanspecifythenameofthequery,whereyouwant
STATISTICAQuerytoputthedatathatthequeryreturns,andadditional
options.
SeetheElectronicManualforfurtherdetails.
IN-PLACE PROCESSING OF
DATA ON REMOTE SERVERS
(THE IDP TECHNOLOGY OPTION)
Thequeryfacilities(describedintheprevioussections),whenofferedaspartof
theenterpriseversionsofSTATISTICA(seeSTATISTICAEnterpriseSystems,page
278),areadditionallyenhancedbyoptionstoprocessdatafromremoteservers
inplace,thatis,withouthavingtoimportthemandcreatealocaldatafile.This
InPlaceDatabaseProcessing(IDP)technologyisparticularlyusefulforprocessing
extremelylargedatafileswhereitcanproducesignificantperformancegainsand
enableSTATISTICAuserstoprocessdatafilesthatexceedthestoragecapacityof
thelocaldeviceoreventheSTATISTICAEnterpriseServer.
Technical note.TheIDPtechnologyisbasedondistributedprocessing
architecture,wherethequeriesareperformedontheserverside(usingtheserver
Chapter9:STATISTICAQuery
PROGRAMMING
STATISTICA
FROM .NET
Adding the STATISTICA Object Library into Your .NET Project .......... 249
Manually Creating the COM Interop Library ....................................... 251
Supporting Multiple Versions of STATISTICA ...................................... 251
Instantiating STATISTICA ...................................................................... 252
The Library Version of STATISTICA ....................................................... 252
1
1
0
0
CHAPTER
1
1
0
0
CHAPTER
Chapter10:Programmingfrom.NET
Atthispoint,thenecessaryCOMInteroplibraryiscreatedautomatically.Under
theprojectReferencesnode,youwillnowseetheentrySTATISTICA.
ThefileInterop.STATISTICA.dllisalsoaddedtotheprojectoutputdirectory.The
STATISTICACOMInteroplibraryisstoredinthisfile.ToviewtheSTATISTICAobject
libraryfromyour.NETproject,rightclickontheSTATISTICAreference,andfrom
theshortcutmenu,selectViewinObjectBrowser.
Chapter10:Programmingfrom.NET
Copyright StatSoft, 2011
STATISTICAQuickReference251
Manually Creating the
COM Interop Library
ItisalsopossibletocreatetheCOMInteroplibrarymanuallyandimportitinto
your.NETproject.Thisgivesyoutheabilitytospecifyadifferentnameforthe
InteropDLLaswellasdefineacustomnamespace.Theprogramthatenablesyou
tocreateanInteropisTLBIMP.EXE.FromaVisualStudiocommandprompt,
executeTLBIMPwithaninitialparameterofthetypelibrarysource.Intheexample
below,theoutputDLLnameandnamespacearealsospecified.
Inthisexample,wereferencethefileSTATIST.EXEsincethatexecutablecontains
theSTATISTICAObjectLibrarytypelibrary.OncetheInteropDLLisgenerated,you
canaddittoyour.NETprojectbyselectingAddReferencefromtheSolution
Explorerasbefore,butthistimeclicktheBrowsebuttontoselectthenewly
createdInteropDLL.
Supporting Multiple Versions
of STATISTICA
TosupportmultipleversionsofSTATISTICA,itisnecessarytomaintainseparate
STATISTICAObjectLibraryInteropDLLsforeachversionofSTATISTICAyouwantto
support.YoucanusetheTLBIMPcommandtogenerateInteropDLLsagainst
specificversionsofSTATIST.EXEandotherDLLs.Whendistributingtheapplication,
ensurethatthecorrectversionoftheSTATISTICAInteropDLLisdeployedwith
your.NETapplication.
Chapter10:Programmingfrom.NET
APPENDIX
A
A
AppendixA:GettingMoreHelp
Thisuniquetextbookhasbeenusedformany
yearsineducationalandresearchactivitiesat
universitiesandresearchorganizations
worldwide.
Other Technical
Support
Resources and
Facilities
Web site resources.StatSofts
Website,oneofthemost
visitedInternetaddresses
relatedtodataanalysis,offers
notonlyaccesstomany
resourcesthatareusefulfor
dataanalysisprofessionalsin
general,butitalsoincludes:
Acontinuouslyupdated
FrequentlyAskedQuestions
section,and
Adownloadareawhereusers
ofthecurrentversionof
STATISTICAproductscan
receivedownloadableupdatesoftheir
software.Weareconstantlyworkingon
increasingthecompatibilityofSTATISTICA
softwareevenwiththoseapplicationsthat
violatestandardconventions.Therefore,in
manycircumstances,downloadinganupdate
canhelpwhentheproblemthatyouare
experiencingiscausedbynonstandardsystem
configurationsorconflictswithother
applications.
E-mail technical support.Ifyourquestionis
notansweredinthelocationsmentioned,you
maysendemailtous.Pleaseincludeyour
serialnumber(inSTATISTICA,selecttheHelp
tab,andintheAboutgroup,clickSTATISTICA
toviewyourserialnumber)andinformation
aboutyourhardware[thetypeofprocessor
(CPU)andtheamountofmemory(RAM)and
diskspace]andtheversionoftheoperating
systemthatyouareusing.
AppendixA:GettingMoreHelp
Copyright StatSoft, 2011
STATISTICAQuickReference259
IfyouliveinNorthAmerica,sendyouremail
toinfo@StatSoft.com;otherwise,emailyour
localStatSoftoffice(seebelow).
Phone technical support.Youcanalsocall
yourlocalStatSoftofficetotalktoa
technician.IfyouliveinNorthAmerica,call
(918)7491119(theNorthAmericantechnical
supportofficehoursare9:00AMto5:00PM
CentralTime,MondaythroughFriday).
Ifyouliveinanotherlocation,pleasecontact
theofficethatservesyourspecificarea.To
locatethatoffice,selecttheHelptab.Inthe
Aboutgroup,clickSTATISTICAtodisplaythe
AboutSTATISTICAdialog,andthenselectthe
InternationalOfficestab.
Pleaseknowyourserialnumber(in
STATISTICA,selecttheHelptab,andinthe
Aboutgroup,clickSTATISTICAtoaccessyour
serialnumber),informationaboutyour
hardware[thetypeofprocessor(CPU)andthe
amountofmemory(RAM)anddiskspace],and
theversionoftheoperatingsystemthatyou
areusingbeforeyoucontactStatSofttechnical
supportoffices.
AppendixA:GettingMoreHelp
STATISTICA ENTERPRISE
SERVER
General Overview ................................................................................... 263
A Broad Choice of Analytic Facilities and Configurations................. 264
Functionality and Applications: The Advantages of
STATISTICA Enterprise Server .......................................................... 264
Advantages of Multithreading Technology ......................................... 265
STATISTICA Enterprise Server User Interface ..................................... 266
Compatibility with Industry Standards ................................................ 269
Architecture of the System (A Technical Note) .................................. 270
Competitive Advantages ....................................................................... 271
Knowledge Portal .................................................................................. 271
STATISTICA Enterprise Server Demo Movie ........................................ 271
B
B
APPENDIX
AppendixA:GettingMoreHelp
Also,whenyoureviewyourSTATISTICA
EnterpriseServeroutputinthebrowser,you
haveoptionstobringanyoralloutputobjects
toyourdesktopcomputerforfurther
processing.Forexample,aclickonasmall
buttonplacedoptionally(dependingontheuser
configuration)nexttoeveryoutputobject
(tableorgraph)senttoyourbrowserbythe
STATISTICAEnterpriseServersystemwilloffer
youtheoptiontodownloadthatobject(a
STATISTICAtableoragraph)totheclient
computerinitsnativeSTATISTICAformat(in.sta
or.stgfileformat)soyoucanworkwithit
offlineusingthelocallyinstalledSTATISTICA
tools.
Advantages of Multithreading
Technology
TheSTATISTICAEnterpriseServerplatformis
builtonadvanceddistributedprocessingand
multithreadingtechnologytosupportoptimal
managementoflargecomputationalloads.This
technologyenablesrapidprocessingofeven
verylargeandcomputationallyintensive
projects,takingfulladvantageofthemultiple
CPUsontheserver,orevenmultipleservers
workinginparallel.
Theillustrationonthenextpageshowsa
projectrunningonaquadprocessorserver,
alongwiththeserverperformancemonitor
demonstratingthefullutilizationofthe
resourcesofallfourCPUsexecutinginthe
multithreadingmodeasingle,computationally
intensiveSTATISTICADataMinerproject.
Inaddition,theSTATISTICAEnterpriseServer
architecturedeliversaplatformindependent,
Webbrowserbaseduserinterface,and
AppendixB:STATISTICAEnterpriseServer
youcanselectadatasource(adatasetoralive
databaseconnection),
reviewandeditthedataintheinteractive
SpreadsheetEditor,
selecttheanalysistobeperformedusingthe
standardmenusystem(orashortcutinthe
userdefinedMyMenu),
selectvariablesandspecifyoptionalanalysis
parameters,
AppendixB:STATISTICAEnterpriseServer
andinteractivelyreviewtheoutput.
Avarietyofinteractivefacilitiestoperform
specialdatabase,qualitycontrol,ordatamining
operations(includinginteractivelybuildingdata
miningmodelsbydraggingarrowsinthemodel
workspace;seebelow)areprovided,andare
accessiblefromthestandardbrowser.
AppendixB:STATISTICAEnterpriseServer
Copyright StatSoft, 2011
STATISTICAQuickReference269
Inadditiontothesebuiltin,straightforward
userinterfacefacilities,STATISTICAEnterprise
Serveralsoincludesatoolkitthatenablesusers
tocustomizetheuserinterfaceanddevelop
customapplicationswithspecificallypredefined
functionality,packagedinawaythatmatches
therequirementsoftheirspecificapplications.
Compatibility with
Industry Standards
Theunsurpassedcompatibilitywithindustry
standardsisanotherinthelonglistofunique
advantagesofSTATISTICAEnterpriseServer.
STATISTICAEnterpriseServercanbedeployed
onanyofthepopularWebserverplatforms
(e.g.,aUNIXbasedApacheorIIS),and
therefore,itwillconformtotheexistinglocal
securityprotocols(firewalls)asrequiredbythe
corporateclient.
STATISTICAEnterpriseServerusesadvanced
proprietarytechnologydevelopedatStatSoftto
ensureitshighperformanceandscalability(e.g.,
multiple,multiprocessorSTATISTICAEnterprise
Servercomputersworkinginadistributed
processingenvironment).Thistechnologyis
builtonStatSoftsyearsofexperienceproviding
highperformance,scalableenterprisesystems
tomajorcorporationsintheUnitedStatesand
aroundtheworld.However,STATISTICA
EnterpriseServerisstillbasedontheindustry
standardcommunicationprotocols(e.g.,XML)
AppendixB:STATISTICAEnterpriseServer
STATISTICA FAMILY
OF PRODUCTS
General Purpose/Desktop Products .................................................... 275
STATISTICA Base .............................................................................. 275
STATISTICA Advanced Linear/Nonlinear Models ........................... 275
STATISTICA Multivariate Exploratory Techniques ......................... 276
STATISTICA Variance Estimation and Precision ............................ 276
STATISTICA Automated Neural Networks (SANN) ......................... 276
STATISTICA Power Analysis ............................................................. 276
Industrial Solutions, Six Sigma Tools .................................................. 276
STATISTICA Quality Control Charts ................................................. 276
STATISTICA Process Analysis .......................................................... 277
STATISTICA Design of Experiments ................................................. 277
STATISTICA Multivariate Statistical Process Control (MSPC) ...... 277
continued
C
C
APPENDIX
STATISTICA Enterprise Systems ........................................................... 278
STATISTICA Data Miner .................................................................... 278
STATISTICA Process Optimization .................................................. 278
STATISTICA Text Miner ..................................................................... 278
STATISTICA Sequence, Association and Link Analysis (SAL) ....... 279
STATISTICA Enterprise ..................................................................... 279
STATISTICA Enterprise/QC .............................................................. 279
STATISTICA Monitoring and Alerting Server (MAS) ....................... 280
STATISTICA ETL (Extract, Transform, and Load) ............................ 280
STATISTICA MultiStream ................................................................. 280
STATISTICA Enterprise Server ......................................................... 281
Scoring Solutions .................................................................................. 281
STATISTICA Live Score ..................................................................... 281
STATISTICA Credit Scoring .............................................................. 281
STATISTICA Scorecard .................................................................... 282
Data and Document Management ...................................................... 282
STATISTICA Document Management System (SDMS) .................. 282
STATISTICA PI Connector................................................................. 283
STATISTICA Data Warehouse .......................................................... 283
Vertical Market Applications ............................................................... 286
PROCEED ........................................................................................... 286
STATISTICA PowerSolutions ............................................................ 287
GENERAL-PURPOSE
DESKTOP PRODUCTS
STATISTICA Base.Offersa
comprehensivesetofessential
statisticsinauserfriendlypackageandallthe
performance,power,andeaseofuseofthe
STATISTICAtechnology.
AllSTATISTICAgraphicstools
BasicStatistics,Breakdowns,andTables
DistributionFitting
MultipleLinearRegression
AnalysisofVariance
Nonparametrics,andmore
STATISTICA Advanced
Linear/Nonlinear Models.Offers
awidearrayofthemostadvancedmodeling
andforecastingtoolsonthemarket,including
automaticmodelselectionfacilitiesand
extensiveinteractivevisualizationtools.
GeneralLinearModels
GeneralizedLinear/NonlinearModels
GeneralRegressionModels
GeneralPartialLeastSquaresModels
NIPALSAlgorithm(PCA/PLS)
VarianceComponents
SurvivalAnalysis
CoxProportionalHazardsModels
APPENDIX
C
C
AppendixC:FamilyofProducts
wasdesignedfor
processindustriesingeneral,butisparticularly
wellsuitedtohelppowergenerationfacilities
leveragetheirdata(collectedintoexisting
specializedprocessdatabasesformultivariate
andpredictiveprocesscontrol)foractionable
advisorysystems.
STATISTICAMultiStreamisacomplete
enterprisesystembuiltonarobust,advanced
clientserver(andfullyWebenabled)
architecture,offerscentraladministrationand
managementofdeploymentofmodels,aswell
ascuttingedgerootcauseanalysisand
predictivedataminingtechnology,andits
analyticsareseamlesslyintegratedwitha
builtindocumentmanagementsystem.
Automated(nonlinear)rootcauseanalysis
andfeatureselectionforthousandsof
parameterstoclearlyidentifywhichones
arethemostlikelyresponsibleforprocess
problems
Automatedandinteractivecommonality
analysistoidentifyparametersand
processesthatshiftedormovedfrom
AppendixC:FamilyofProducts
Copyright StatSoft, 2011
STATISTICAQuickReference281
normaloperationsduringparticulartime
intervals
Advancedlinearandnonlinear(e.g.,SVM,
RecursivePartitioning,NeuralNets)models
forcreatingsensitivemultivariatecontrol
schemesandworkflowstoidentify
multivariateshiftsanddriftsearly,before
theycauseproblems
Advanceddataminingalgorithmsfor
predictingandoptimizingkeyperformance
andqualityindicators
Trackshundredsofdatastreams
simultaneously
Deliverssimplesummariesrelevantto
criticalprocessparametersandoutcomes
viaefficientandsimpledashboardsand
drilldownworkflows
Deliversstandardandcustomizedanalytic
workflowsforrootcauseanalysis,
leveragingcuttingedgedataanalysisand
dataminingtechnologies
Warnsof(predicted)problemsand
equipmentfailuresbeforetheyoccur
(predictivealarming),thusavoidingcostly
shutdownsandunscheduledmaintenance
Watcheseverythingthatimpactsyour
processperformanceinrealtime
STATISTICA Enterprise Server.The
ultimateenterprisesystemthatoffers
fullWebenablement,includingtheabilityto
runSTATISTICAinteractivelyorinbatchfroma
Webbrowseronanycomputer(including
Linux,UNIX)andoffloadtimeconsumingtasks
totheservers(usingdistributedprocessing).
UsesmultitierClientServerarchitecture,
supportingmultithreadinganddistributed/
parallelprocessingthatscalestomultiple
servercomputers.
SCORING SOLUTIONS
STATISTICA Live Score.STATISTICA
EnterpriseServersoftwarewithinthe
STATISTICADataAnalysisandDataMining
Platform.Dataareaggregatedandcleaned
andmodelsaretrainedandvalidatedusingthe
STATISTICADataMinersoftware.Oncethe
modelsarevalidated,theyaredeployedtothe
STATISTICALiveScoreserver.STATISTICALive
Scoreprovidesmultithreaded,efficient,and
platformindependentscoringofdatafrom
lineofbusinessapplications.Someexamples
oftheuseofSTATISTICALiveScore:
Providescreditscorecardstocustomer
serviceapplications(e.g.,callcentersystems
andWebbasedapplications)
Enablescustomersegmentation,up
sell/crosssell,andcustomerchurn
identificationtocustomerserviceand
marketingrepresentatives
Providesproactivefrauddetectionalertsto
analysts
STATISTICA Credit Scoring.The
solutionforanycompanytobuildin
housemodelsforitsvariouscreditproducts
anddecisionmaking.STATISTICACreditScoring
coversallaspectsofthecreditscoringneeds
foryourcompany.
In-house model building.TheSTATISTICA
CreditScoringsoftwaresolutionenablesthe
developmentandevaluationofpredictive
modelstoevaluateandassignariskto
applicationsforcredit,eitherforarequest
foranewaccountorforrequestedchanges
(e.g.,balanceincrease)tothetermsofan
existingcreditaccount.
Scoring applications.STATISTICALive
Scoreenablescompaniestoscorecredit
applications;itcanbeeasilyintegratedwith
yourexistingcustomerservicesystems,self
serviceWebsitesforcustomers,etc.
AppendixC:FamilyofProducts
QuickReference:Index
Copyright StatSoft, 2011
STATISTICAQuickReference289
INDEX
A
accept/rejectattribute,55
Acrobatreports,153
ActiveX,169,181,198,238
documents,238
objects,238
adhocbygroupanalyses,50
advancedlinear/nonlinear
models,275
Advancedtab,18
advice,statistical,33
aggregation,93
AIAGMSAmanual,55
AllSpecsbutton,14
analyses
attributegage,55
automating,40
autominimize,129
buttons,analysisbar,129,
135
closeall,130,137
manufacturing,55
quickvs.advanced,18
recording,230
rerun,236
resume,38,237
selection,16,17
analysisbar,129,130,135
analysisconfiguration,
STATISTICAEnterprise,120
analysismacros,224
analysisspecificationdialogs,
131
analysissummary,54
analysisworkbooks,22
Analysis/GraphOutput
Managerdialog,23,25
analyticfacilities,3
analyticsexamples,11
analyzinglargedata
problems,50
annotations,149
ANOVA
example,34
onewaydesigns,34
repeatedmeasuresdesigns,
34
appendsupplementary
information,132
applicationobject,252
arrangementoffactors,39
attributegageanalysis,55
audittraillogging,103
audittrail,spreadsheets,106
autofiltering,133
autosave,148
automatedneuralnetworks,
276
B
batchformulas,72,75
BFGSalgorithm,276
blockdatagraphs,199,202
block,deselect,17
brushing,132,205
Brushingdialog,205
bundles,variable,40
buttons
AllSpecs,14
ByGroup,44
Functions,74
OK,19
OpenData,13
Options,23,25,134
Spread,20
Summary,19
Variables,19
Zoom,20
ByGroupbutton,44
bygroupanalyses,47
example,43
C
C/C++,59,227,276
canonicalanalysis,276
capabilityanalysis.See
processcapabilityanalysis
caseheaders,175
caselabels,207
casestates,132,205
excluded,207
hidden,207
cases
filterduplicates,85
causeandeffectdiagrams,
277
cellformatting,spreadsheets,
176
centralcompositedesigns,
277
classicmenus,11,12
classificationtrees,276
cleaningdata,84
closeallanalyses,130,137
closeallwindows,137
clusteranalysis,276
codes,36,109
missingdata,90
COMInteroplibrary,251
compliancerequirements,
meeting,105
configurations,different,218
configurations,network,218
conjugategradientalgorithm,
276
copy,23
copywithheaders,23
correlationmatrix,16
correlationsexample,11
correlations,significant,21
correspondenceanalysis,276
Coxproportionalhazards
models,275
creationstamp,109
QuickReference:Index
dialogs
Analysis/GraphOutput
Manager,23
analysisspecification,131
Analysis/GraphOutput
Manager,25
autominimize,136
Brushing,205
Customize,139
DatabaseConnection,80
FunctionBrowser,74
OpenaSTATISTICAData
File,13
Options,15,25,134,215
outputselection(results),
132
PrintSpreadsheet,24
results,132
selfprompting,19
StartupPanel,13,131
UserInterface,11
VariableBundleManager,
40
variableselection,19,133
variablespecifications,13
VariableSpecifications
Editor,14
WelcometoSTATISTICA,12
DIN55319,52
discriminantanalysis,276
distributionmodel,time
dependent,54
documentcustomization,214
documentmanagement
system,163,282
documenttypes,137
documents,recentlyused,
138
draganddrop,182
QuickReference:Index
Copyright StatSoft, 2011
STATISTICAQuickReference291
E
Edittab,29
ElectronicManual,26,33,36,
257
ElectronicStatisticsTextbook,
27,258
EnhancedSVB,222
Enterpriseinstallations,98
enterprisenetwork,279
enterprisesystems,278
enterprise/QCnetworks,279
EWMAchart,277
exampledatasets,45
examples
accessingdatadirectlyfrom
databases,79
analytics,11
ANOVA,34
bygroupanalyses,43
correlations,11
datapreparationcleaning
andfiltering,84
getexternaldatavia
STATISTICAQuery,244
inputdatadirectlyfrom
Excel,77
macrorecording,230
recordingananalysis,230
spreadsheetformulas,
batchformulas,72
STATISTICADataMiner
Recipes,63
STATISTICAEnterprise,109
STATISTICAEnterprise
Server,98
STATISTICAVisualBasic,
230
summaryresultspanels,51
usingSTATISTICAExtract,
TransformandLoad,93
examples(cont.)
usingSTATISTICAin
regulatedenvironments,
102
variablebundles,40
Excel,77,140,142,148,151,
169,180,182,198,238
inputdatadirectlyfrom,77
openinSTATISTICA,142
exploratorydataanalysis,44,
50
exportoutput,7
extract,transform,andload,
280
F
F1key,13
factoranalysis,276
factors,arrangement,39
filterdata.Seedatacleaning
andfiltering
filterduplicatecases,85
filtersparsedata,87
filteringvariables,133
fixednonlinearregression,
276
formulaeditor,72
formulas,14,72
multiple,75
results,73
spreadsheet,14
fractionalfactorialdesigns,
277
frauddetection,281
frequencytables,51
fromclauses,STATISTICA
Query,180
function
externallycallable,4,227,
228,275
internallyused,12,73,74,
104,196,226
FunctionBrowserdialog,74
Functionsbutton,74
G
gagerepeatability/
reproducibility,277
generaldiscriminantanalysis
models,276
generallinearmodels,36,275
generaloverview
analyticfacilities,3
softwaretechnology,6
uniquefeatures,4
Webenablement,7
generalpartialleastsquares
models,275
generalregressionmodels,275
generalizedlinear/nonlinear
models,275
globalmacros,228
gradientdescentalgorithm,
276
graphs,182,189
autoupdating,143
blockdata,199,202
brushing,205
casestates,205
categories,198
creatingviaSTATISTICA
VisualBasic,209
custom,203
customization,29,190,217
customizing,203
default,203
defaults,217
drawingtools,29
inputdata,198,199
piecharts,194
producedfrom
spreadsheets,28
shortcutmenus,29
specialized,208
QuickReference:Index
variable
block,17,19
changeformat,14
changename,14
formula,14
processinvariant,88
selection,19
selectionconventions,19
specifications,13
VariableBundleManager
dialog,40
variablebundlesexample,40
Variablesbutton,19
variableheaders,176
variableselectiondialog,133
variablespecificationsdialog,
13
VariableSpecificationsEditor,
14
variables
automaticprescreening,
133
bundles,40
ToolTips,43
filtering,133
measurementtypes,133
organizelargesets,40
reorder,47
repeatedselection,40
variancecomponents,275
variancecomponentsfor
randomeffects,277
varianceestimationand
precision,276
Viewtab,215
VisualBasic,221
methods,141
properties,141
W
webbrowser,usingwith
STATISTICA,98
webenablement,7
weboutput,16,155
website,StatSoft,258
Weibullanalysis,277
WelcometoSTATISTICA
dialog,12
whereclauses,STATISTICA
Query,180
Word,140,142,143,148,
154,169,182,198,238
workbooks,22,148,169
draganddrop,172
icons,172
notesandcomments,149
overview,169
printdocumentfrom
within,24
redarrow,236
rerunninganalyses,236
saveaswebpages,150
tabs,138,170
tree,171
X
XbarandRcharts,277
XML,278
Z
Zoombutton,20
QuickReference
Copyright StatSoft, 2011
STATISTICAQuickReference297
QuickReference