單方程第3題分位數(shù)回歸模型例4.10分位數(shù)回歸

1、單方程第3題分位數(shù)回歸模型-例4.10分位數(shù)回歸分位數(shù)回歸建立如下的回歸方程研究政府支出對(duì)居民消費(fèi)的影響：ln(cs)，卩+卩l(xiāng)n(inc)+卩l(xiāng)n(cs)+卩l(xiāng)n(fe)t01t2t13t1首先進(jìn)行數(shù)據(jù)處理(定義數(shù)據(jù)庫(kù)里沒有的時(shí)間序列數(shù)據(jù))：?jiǎn)螕鬛uick單擊Generateseries分別輸入：PlEViews-Workfile:徐菌例410-Cc:documentsandsettingsadministratorFileEditObjectViewPoizQuickOptionsAdd-inrWindowHelpViewProcObjectPrintSaveDetaikSample.rS

2、ampleRange:19782006-29obsSample:19782006-29obsGeneraiteSeries.Show.Graph.EmptyGroup(EditSeriesSeriesStatistic:卜GroLi_pStatistic&卜EstimateEquation.EstimateVAR,.tcpics_rcs_ucspexpfincprpuresidydrcsp=(cs_u*pu+cs_r*pr)/cpiGenerateSeriesbyEquationr-Enterequationcsp=(c5_u+：pu+c5_r+pr)/cpi|r-53mpl&-1978200

3、6OK1Cancelinc=(yd_r*pr+yd_u*pu)/cpi下面先進(jìn)行OLS估計(jì)OLS估計(jì)：操作步驟：方法一：在命令窗口輸入：lsLOG(CSP)cLOG(INC)LOG(CSP(-1)LOG(EXPFP(-1)回車即得到下圖：0EViewsFileEditObjectViewProcQuickOption:Ackl-in:WindowHelpIsLOG(CSP)cLOG(INC)LOG(CSPC-I)LOG(EXPFP(-1)THWorkfile:徐菌例410-c:dacumentsandsettingsadministrat.-xviewprodotjjEctPrint#Ecju

4、mtiari：UNTTTLEDWorkfilE：徐菌例410:UntitlEcl-XRange:19732006-2!Sample:197S2006-prQCObjectPrintNameViewFreezeEstimateForecastStatsResidscpics_rcs_ucspexpfe：!pfpincpr卩iresidDependentVariable:LOG(CSP)Method:LeastSquaresDate:01/06/14Time:11:42Sample(adjusted):19792006Includedobservations:28afteradjustmentsa

5、iIInt社led&PageVariableCoefficientStd.Errort-StatisticProb.CO.202S240.04S9505.7770610.0000LOG(INC)0.4660890.0645647.2139720.0000LOG(CSP(-1)0.4705840.0621717.5691720.0000LOG(EXPFP(-1)0.0267580.0161941.6523570.1115R-squared0.999283Meandependentvar0.536631AdjustedR-squared0.999198S.D.dependentvar0.60439

6、6S.E.ofregression0.017111Akaikeinfocriterion-5.166570Sumsquaredresid0.007027Schwarzcriterion-4.976263Loglikelihood76.33210Hannan-Quinncriter.-5.100397F-statistic11220.30Durbin-Watsonstat1.957044Prob(F-statistic)0.000000方法二：利用菜單操作0EViews-Workfile:徐茵例41。-(c:documentsandsettingsadministratorSFileEditOb

7、jectViewProcQuickOptioniEAclcl-iri5WindowHelpViewProcObjectPrintSave|Detail?Saftple.rSampleRange:19782006-29obsGenerate5亡ries.Sannple:19782006-29obs宜m.Ic0cpiGmph.0c右0CS_LEmptyGroupEditSeries-)0csp0卩0e?(pfp0incSprSeriesStatiticGrouStatic匚$EstimateEquation.0PJEstimateVAR.0resid回心0yd_uSEViews-WorkHe:徐苗

8、例410(cdocumentsandsetting血dmnistrata諄面討童經(jīng)濟(jì)學(xué)上機(jī)考試i計(jì)量蛭濟(jì)廿祐方法與建模習(xí)題】H囹FileEditObjectViewProcQuickOptionsAdd-injWindowHelpIVelvPratObjectPrintanr亡Details*/-Feich5-icreDeleteGenrSampleRange:19702006-Q9absSample.19T8200B-29QbG回0000000000回s叩i吒匸s_uL：SflBCpf跑殆meprpuresidEquationEstimationxSpacficaHcnQvhzm保存回歸的方

9、程：點(diǎn)擊name，取名為:ep_olsEViews-Equation:EWorkfile:4L0:UntitledITTllFileEditObjectViewProcQuickOptionsAdd-insWindowHelpDopendentVariable:LOG(CSPJMethod:LeastSquaresDate:01/06/14Time:11:45Sample(adjusted):197920D6Includedobservaiions:28afteradjustmentsObjectNameXvariablecoeincien:5td.Errort-StatlSticFrob.0

10、0.2023240.0480505.7770610.0000LOG(INC)0.4660390.0645647.2109720.0000L0G(CSP(-1)0.4J05S40.0621717.5691720.0000LOG(EXPFP(-1)0.0267580.0161941.6523570.1115R-squared0.999203Meandependentvar3.536631AdjustedR-squared0999198SDdependentvar0.604396S.E.ofregression0.017111Akaikeinfocriterion-5.16657SGurnsquar

11、edreeid0.007027Schwarzcriterion-4.9762B3LoglikBlihuiid76.33210HannarvQuinncriter.-6.108307F-statistic11220.30Durbin-Watsonstat1.957044Prob(F-statistic)i：i.i：ii：ii：iii：ii：i2斗匚hsractersmoKimumi16orFewerrecommeridsdCispiayn-smeforlabelingtablesandgraphs(optionali-WarnetoidentiF*objectCancelvieirtProcOb

12、jectPrintName尸reezeEstimateForecast5tatsEViews-Equation:EWorkfile:4L0:UntitledITTllFileEditObjectViewProcQuickOptionsAdd-insWindowHelpEViews-Equation:EWorkfile:4L0:UntitledITTllFileEditObjectViewProcQuickOptionsAdd-insWindowHelp分位數(shù)回歸分析操作步驟：分位數(shù)回歸估計(jì)中，點(diǎn)擊Quick-Estimateequation-在equationspecification框里輸入

13、LOG(CSP)cLOG(INC)LOG(CSP(-1)LOG(EXPFP(-1)，選擇QREG方法，并在“quantitleto”填入相應(yīng)的分位數(shù)，如0.2記住，每次得到的結(jié)果都要保存一下哦！EViews-Equation:EWorkfile:4L0:UntitledITTllFileEditObjectViewProcQuickOptionsAdd-insWindowHelpEViews-Equation:EWorkfile:4L0:UntitledITTllFileEditObjectViewProcQuickOptionsAdd-insWindowHelp分位數(shù)為0.2：0EViews

14、-Equation:EQ_QREG02Workfile:徐茴例41Q:UntitledFileEditOtyectViewProcQuickOption:Add-in:Window卜viewProcObjectPrintNameFreezeEstimateForecastStatsReaidsDependentvariable:LOG(CSP)Method:QuantileRegressionftau=0.2)Date:01/06H4Time:11:52Sample(adjusled):19792006Includedobservations:28afteradjustmentsHuberSa

15、ndwichStandardErrors&CovarianceSparsitymethod:KernelEpanechnikov)usingresidualsBandwidthmethod:Hall-Sheather,bvO.18828EstimaiionsuccessfullyidentifiesuniqueoptimalsolutionVariableCoefficientStd.Errort-StatisticProb.C0.20955S0.0752572.7S45470.0103LOGCINC)0.4S53830.1082084.4856300.0002LOG(CSP(-1)0.43S

16、1560.1039404.2154700.0003LOG(EXPFPf-IO)0.0475700.0293631.6200650.1183PseudoR-squared0.972919Meandependentvar8.536631AdjustedR-squared0.9695348.D.dependentvar0.604396S.E.cfregression0.029215Objective0.125904Quantiledependentvar7.990470Restr.objective4.649230Sparsity0.004695Quasi-LRstatistic667.5910J?

17、rob(Quasi-LRstat)0.000000分位數(shù)為05：0EViews-Equation:UNTITLEDWorkfile:徐菌例410:UntitledY|FitEditObjettViewProcQuickOption:Adcl-in:WindowI-viewProcObjectPrintNameFreezeEstimateForetsstStatsResidsDeperdentVariable:LOGCCSP)Method:QuantileRegression(Median)Date:O1/oe/14Time:11:53Sample(adjusted):19792006Inclu

18、dedobservations:2Safteradjustment?HuberSandwichStandardErrors&CovarianceSparsitymethod:Kernel(Epanechnikov)usirgresidualsBandwidthmethod:Hall-Slieather,bvO.31995EstimationsuccessfullyidentifiesuniqueoptimalsolutionVariableCoefficientStd.Errort-SiatisticProb.C0.2520090.033573.43536S0.0022LOG(INC)0.37

19、S5650.1326622.3273130.02S7LOG(CSP(-10.5585590.1575003.5464100.0016LOG(EXPFP(-1)0.0335300.02S0391.195S310.2434PseudoR-squared0.974097MeandependentvarS.536631AdjusledR-squared0.970859S.D.血卩endentvar0.BO439BS.E.ofregression0.017829Objetiive0.1S07E9Quantiledependentvar3.436459Restr.otojeclive6.979443Spa

20、rsiiy0.054510Quasi-LRstatistic997.78E7frQb(Guasi-LRstat)0.000000分位數(shù)為010EViews-Equation:UNTITLEDWorkfile;徐苗例410;UntitledFileEditObjectViewProcQuickOptioriEAtltl-irirWirclowI-(viewProcObjectPrintNameFreezeEstimateForecastitatsResitisDependentVariable:LOG(CSP)Method:QuantileRegression(tau=0.9)Date:01/0

21、6/14Time:11:54Samplefadjustedy19792006Includedobservalions:28afteradjustmentsHuberSandwichStandardErrorsCovarianceSparsitymethod:Kernel(Epanethnikov)usingresidualsBandwidthmethad:Hall-Sheather,bwO.18828EstimationsuccessfullyidentifiesuniqueoptinralsolutionVariableCoefficientSid.Errort-StaiisticProb.

22、C0.2776540.0876643.1672460.0042LOGflNC0.4461840.1521952.9316680.0073LOG(CSP(-1)0.4945770.1440413.43357S0.0022LOG(EXPFP(-1)J0.0259010.0316730.8177420.4215PseudoR-squared0.976689Meandependentvar8.536631Adjust已dR-sqjar已d0.973775S.D.deoeridentvar0.604396S.E.ofregressionC.025339Objective0.111331Quantiledependentvar9.105995Restr.ctijective4.775853Sparsity