[519993基金]Spearman's rank correlation coefficient (斯皮尔曼秩相关系数)
作者
Spearman'srankcorrelationcoefficientFromWikipedia,thefreeencyclopediaJumpto:navigation,searchInstatistics,Spearman'srankcorrelationcoefficientorSpearman'srho,namedafterCharlesSpearmanandoftendenotedbytheGreekletterρ(rho)orasrs,isanon-parametricmeasureofcorrelation–thatis,itassesseshowwellanarbitrarymonotonicfunctioncoulddescribetherelationshipbetweentwovariables,withoutmakinganyassumptionsaboutthefrequencydistributionofthevariables.Contents1Calculation2Example3Determiningsignificance4CorrespondenceanalysisbasedonSpearman'srho5Seealso6Notes7References8ExternallinksCalculationInprinciple,ρissimplyaspecialcaseofthePearsonproduct-momentcoefficientinwhichtwosetsofdataXiandYiareconvertedtorankingsxiandyibeforecalculatingthecoefficient.[1]Inpractice,however,asimplerprocedureisnormallyusedtocalculateρ.Therawscoresareconvertedtoranks,andthedifferencesdibetweentheranksofeachobservationonthetwovariablesarecalculated.Iftherearenotiedranks,i.e.thenρisgivenby:where:di=xi?yi=thedifferencebetweentheranksofcorrespondingvaluesXiandYi,andn=thenumberofvaluesineachdataset(sameforbothsets).Iftiedranksexist,classicPearson'scorrelationcoefficientbetweenrankshastobeusedinsteadofthisformula:[1]Onehastoassignthesameranktoeachoftheequalvalues.Itisanaverageoftheirpositionsintheascendingorderofthevalues:AnexampleofaveragingranksInthetablebelow,noticehowtherankofvaluesthatarethesameisthemeanofwhattheirrankswouldotherwisebe.VariableXi?Positioninthedescendingorder?Rankxi0.8?5?51.2?4?1.2?3?2.3?2?218?1?1Inthiscasewecannotusetheshortcutformula(becauseofthetiedranksinthedata)andmustusethesecond,product-momentform.ExampleTherawdatausedinthisexampleisshownbelowwherewewanttocalculatethecorrelationbetweentheIQofsomeonewiththenumberofhoursspentinfrontofTVperweek.IQ,Xi?HoursofTVperweek,Yi106?786?0100?27101?5099?28103?2997?20113?12112?6110?17Thefirststepistosortthisdatabythefirstcolumn.Next,twomorecolumnsarecreated(xiandyi).Thelastofthesecolumns(yi)isassigned1,2,3,...n,andthenthedataissortedbythefirstoriginalcolumn(Xi).Thefirstofthenewlycreatedcolumns(xi)isassigned1,2,3,...n.Thenacolumndiiscreatedtoholdthedifferencesbetweenthetworankcolumns(xiandyi).Finallyanothercolumnshouldbecreated.Thisisjustcolumndisquared.Afterdoingthisprocesswiththeexampledatayoushouldendupwithsomethinglike:IQ,Xi?HoursofTVperweek,Yi?rankxi?rankyi?di?86?0?1?1?0?097?20?2?6?-4?1699?28?3?8?-5?25100?27?4?7?-3?9101?50?5?10?-5?25103?29?6?9?-3?9106?7?7?3?4?16110?17?8?5?3?9112?6?9?2?7?49113?12?10?4?6?36Thevaluesinthecolumncannowbeaddedtofind.Thevalueofnis10.Sothesevaluescannowbesubstitutedbackintotheequation,whichevaluatestoρ=?0.175758whichshowsthatthecorrelationbetweenIQandhourspendbetweenTVisreallylow(barelyanycorrelation).Inthecaseoftiesintheoriginalvalues,thisformulashouldnotbeused.Instead,thePearsoncorrelationcoefficientshouldbecalculatedontheranks(wheretiesaregivenranks,asdescribedabove).DeterminingsignificanceThemodernapproachtotestingwhetheranobservedvalueofρissignificantlydifferentfromzero(wewillalwayshave1≥ρ≥?1)istocalculatetheprobabilitythatitwouldbegreaterthanorequaltotheobservedρ,giventhenullhypothesis,byusingapermutationtest.Thisapproachisalmostalwayssuperiortotraditionalmethods,unlessthedatasetissolargethatcomputingpowerisnotsufficienttogeneratepermutations,orunlessanalgorithmforcreatingpermutationsthatarelogicalunderthenullhypothesisisdifficulttodevisefortheparticularcase(butusuallythesealgorithmsarestraightforward).Althoughthepermutationtestisoftentrivialtoperformforanyonewithcomputingresourcesandprogrammingexperience,traditionalmethodsfordeterminingsignificancearestillwidelyused.Themostbasicapproachistocomparetheobservedρwithpublishedtablesforvariouslevelsofsignificance.Thisisasimplesolutionifthesignificanceonlyneedstobeknownwithinacertainrangeorlessthanacertainvalue,aslongastablesareavailablethatspecifythedesiredranges.Areferencetosuchatableisgivenbelow.However,generatingthesetablesiscomputationallyintensiveandcomplicatedmathematicaltrickshavebeenusedovertheyearstogeneratetablesforlargerandlargersamplesizes,soitisnotpracticalformostpeopletoextendexistingtables.AnalternativeapproachavailableforsufficientlylargesamplesizesisanapproximationtotheStudent'st-distributionwithdegreesoffreedomN-2.Forsamplesizesaboveabout20,thevariablehasaStudent'st-distributioninthenullcase(zerocorrelation).Inthenon-nullcase(i.e.totestwhetheranobservedρissignificantlydifferentfromatheoreticalvalue,orwhethertwoobservedρsdiffersignificantly)testsaremuchlesspowerful,thoughthet-distributioncanagainbeused.AgeneralizationoftheSpearmancoefficientisusefulinthesituationwheretherearethreeormoreconditions,anumberofsubjectsareallobservedineachofthem,andwepredictthattheobservationswillhaveaparticularorder.Forexample,anumberofsubjectsmighteachbegiventhreetrialsatthesametask,andwepredictthatperformancewillimprovefromtrialtotrial.AtestofthesignificanceofthetrendbetweenconditionsinthissituationwasdevelopedbyE.B.PageandisusuallyreferredtoasPage'strendtestfororderedalternatives.CorrespondenceanalysisbasedonSpearman'srhoClassiccorrespondenceanalysisisastatisticalmethodwhichgivesascoretoeveryvalueoftwonominalvariables,inthiswaythatPearson'scorrelationcoefficientbetweenthemismaximized.Thereexistsanequivalentofthismethod,calledgradecorrespondenceanalysis,whichmaximizesSpearman'srhoorKendall'stau[2].SeealsoStatisticsportalKendalltaurankcorrelationcoefficientRankcorrelationChebyshev'ssuminequality,rearrangementinequality(ThesetwoarticlesmayshedlightonthemathematicalpropertiesofSpearman'sρ.)Pearsonproduct-momentcorrelationcoefficient,asimilarcorrelationmethodthatinsteadreliesonthedatabeinglinearlycorrelated.Notes^abMyers,JeromeL.;ArnoldD.Well(2003).ResearchDesignandStatisticalAnalysis,secondedition,LawrenceErlbaum,p.508.ISBN0805840370.^Kowalczyk,T.;PleszczyńskaE.,RulandF.(eds.)(2004).GradeModelsandMethodsforDataAnalysiswithApplicationsfortheAnalysisofDataPopulations,StudiesinFuzzinessandSoftComputingvol.151.BerlinHeidelbergNewYork:SpringerVerlag.ISBN9783540211204.References
金融工程,数学算法,the,ranks原文发布于宽客论坛,点击阅览原文
金融工程,数学算法,the,ranks原文发布于宽客论坛,点击阅览原文
目录
推荐阅读
-
封闭式基金价格由(博时增长前基金)
记者黄一灵见习记者石诗语“实在是意料之外,我8月中旬开端走的商贷借款批阅,其时告知我至少要比及下一年1月,没想到本年10月中旬借款...
-
中国中小学联率达10期货单边去高值0%
现在,我国中小学(含教育点)联网率到达100%,出口带宽到达100兆的校园占99.9%,99.5%的中小校园具有多媒体教室,教育信...
-
恒丰银行定增1000亿股 筹集资金1000亿 中央汇金认期货配资鑫茂杭州购6成
恒丰银行重组有了新进展!12月18日,恒丰银行宣告非公开发行1000亿普通股股份,每股价格1元,其间,中心汇金拟认购600亿股,山...
-
[不同的股票指数]买基金是怎么赚取利润
基金盈亏办理人员都拿酬劳,分奖金,**不赔,可那是买基金人的钱,基金办理人搞利益输送的,买基金的人就更惨了!绝不让基金办理人割韭菜...
-
[第一财经 头脑风暴]生孩子新农合可以报销吗 需要哪些资料
现在大部分人成婚之后都是想要一个归于自己的宝宝,哺育下一代是每一辈人的严重方针。在这整个过程中,生孩子是比较困难的,是需求花费...
-
怎么样买杨天南的基金(交银蓝筹基金今日净值)交银蓝筹基金今日净值
大智慧阿思达克通讯社10月21日讯,九鼎出资(430719)20日公告称,已于近来与天源证券有限公司各股东签署协议,公司出资363...
-
[智冠]股票期权行权
国信证券开户的能申购国信证券新股吗???每一个申购单位为1000股,申购数量应当为1000股或其整数倍,但最高不得超越当次网上初始...