eviews的相关计量经济学操作_第1页
eviews的相关计量经济学操作_第2页
eviews的相关计量经济学操作_第3页
eviews的相关计量经济学操作_第4页
eviews的相关计量经济学操作_第5页
已阅读5页,还剩5页未读 继续免费阅读

下载本文档

版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领

文档简介

Wew/运用于计童霾济学的裁个拳逵检验:首先做出相应的Is模型,□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\叵S3View|Proc|Object|Print|Name|Freeze|EstimateForecast|Stats|ResidsDependentVariable:YMethod:LeastSquaresDate:10/17/12Time:19:45Sample:19912010Includedobservations:20VariableCoefficientStd.Errort-StatisticF「Ot)・CX272.22500755152159.64391.7052010.02331132.394510.10540.0000R-squaredAdjustedR-squaredS.E.ofregressionSumsquaredresidLoglikelihoodF-statisticProb(F-statistic)0.9831370.982200216.8400846352.2-134.90841049.4040.000000MeandependentvarS.D.dependentvarAkaikeinfocriterionSchwarzcriterionHannan^Quinncriter.Durbin-Watsonstat5199.5151625.27513.6908413.7904113.710271.365194White检验:在HeteroskedasticityTest:White中检验p值,如果p・f值小T0.05则表示有异方星反之没有异方差。

23HeteroskedasticityTests23HeteroskedasticityTests□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\|~^~||~^序View|Proc|Object|Print|Name|Freeze|EstimateForecast|StatsResids|Heteroskedasticity*Test:WhiteF-statistic14.62196Prob.F(2,17)0.0002Obs^R-squared12.64768Prob.Chi-Square(2)0.0018ScaledexplainedSS5.563250Prob.Chi-Square(2)0.0619\_ITestEquation:DependentVariable:RESIDEMethod:LeastSquaresDate:10/17/12Time:19:46Sample:19912010Includedobservations:20VariableCoefficientStd.Errort-StatisticProb.C-180508.7103270.4-1.7479220.0985X49.2939828.925851.7041500.1066XA2-0.0021070.001846-1.1410190.2697R-squared0.632384Meandependent回42317.61AdjustedR-squared0.589135S.D.dependent炖45247.09S.E.ofregression29002.81Akaikeinfocriterion23.52565Sumsquaredresid1.43E*10Schwarzcriterion23.67501G—Q检验20对数据中在上方输入sortxsmpl17Isyex承排序承选出1-7承做回归得出Sumsquaredresid同样再输入smpl1420/lsycx得出第二个rss用rss⑴/rss(2)做f检验其他检验方式中用genr定义变量进行回归分析确定最人的r2等值来确定其异方差形式修正方法加权最小2定义el为相应的resid(e)在Is规划时将opion中的weight设为abs(el)进行来说规划即可这时会去掉其异方差性对数据建模Is分析数据看哪个每个解释变量的f值,检验其是否有多重共线性求其相关系数矩阵“Quick'GroupStatistics'Correlations"得出后和0.8比较,如果大于0.8说明两者之间有激情。修正方法:

逐步回归先对每一个进行Is分析建模然后取出对y影响最人的做为基础然后更具其相关系数人小排序,用EquationEstimation做出先关的检验,选择加入的元素EquationEstimation做出先关的检验,选择加入的元素口Equation;UNTITLEDWorkfile;UNTITLED;;Untitled\口Equation;UNTITLEDWorkfile;UNTITLED;;Untitled\|a||E)||S3|口Equation;UNTITLEDWorkfile;UNTITLED;;Untitled\口Equation;UNTITLEDWorkfile;UNTITLED;;Untitled\|a||E)||S3|View|Proc|Object|Print|NameFreeze|EstimateForecastStatsResids|DependentVariable:YMethod:StepwiseRegressionDate:10/17/12Time:19:04Sample:19741987Includedobservations:14Numberofalwaysineludedregressors:2Numberofsearchregressors:4Selectionmethod:StepwiseforwardsStoppingcriterion:p-valueforwards/backwards=0.5/0.5VariableCoefficientStd.Errort-StatisticProb*C-39.7947925.01570-1.5907930.1400X10.2115430.0453024.6695810.0007X31.9092460.7241532.6365230.0231R-squared0.95394-5Meandependerrt炖142.7129AdjustedR-squared0.945571S.D.depend©nt回26.09805S.E.ofregression6.088671Akaikeinfocriterion6.638146Sumsquaredresid407.7910Schwarzeriterion6.775087Loglikelihood-43.46702HmnnNn-Quinncriter.6.625470「statistic113.9220Durbin-Watsonstat1.655554Prob(「statistic)0.000000SelectionSummaryAddedX3^Note:p-values汕dsubsequenitesisdonotaccountfoestepwiseselection.如上图就是加入了x3同样的Is对其d.w做出分析,如呆接近2则没有一次相关,若出了范闱则有相应的相关性其他次的相关性可以由Im检验得到(小于0.05即有序列相关性)。自相关的检验还有view/residual/con——Q-——做出如图所示的表Date:10/17/12Time:20:42Sample:19902009Ineludedobservations:20ACPAGQ-StatProbAutocorrelationPmrtialCorrelationACPAGQ-StatProb10.2130.21310.2130.2131.04830.30620.039-0.0071.08480.5813-0.058-0.0681仃130.7604-0.029-0.002119370.8795-0.235-0.2372.80870.7296-0.275-0.2025.19040.5207-0.231-0.1566.99060.4308-0.186-0.1798.25850.40990.0750.1088.48120.48610-0.069-0.2088.69240.562110.056-0.0338.84550.63612-0.039-0.2108.92890.709修正方法杜宾两步法:进行如卜的Is估计Z=0o(l_Q)+0i(X厂QX—J+%+修正方法杜宾两步法:进行如卜的Is估计Z=0o(l_Q)+0i(X厂QX—J+%+片EquationEstimationI22ISpecificationOptions]Equationspeci£icationDependentvariablefoilowedbylistofregressorsandPDLterms;ORexplicitq.uationlik«elog(y)cLogty(-1))logCx)log(x(-1))Estim^tioxisettingsththod:~~-~~LeastSquares~ftILSandARMA)▼|Sample:119802000~确定取消可得:□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\|~^||~^|□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\|~^||~^|匡□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\\\~W||~^□Equation:UNTITLEDWorkfile:UNTITLED::Untitled\\\~W||~^View|P「x|Qbj^c:t|Prin11Name:|Freeze:|EstimateForEGast|Stats|R亡sidsDependentVariable:LOG(Y)Method:LeastSquaresDate:10/17/12Time:19:27Sample(adjusted):19812000Includedobservations:20afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.C0.4568710.1570642.9088170.0103LOG(Y(-1))0.6278730.0874307.1814730.0000LOG(X)0.4657910.0799945.8228360.0000LOG(X(-1))-0.1247940.116974-1.0668570.3019R-squared0.998506Meandependentvar9.102328AdjustedR-squared0.998226S.D.dependentvar1.035997S.E.ofregression0.043633Akaikeinfocriterion-3.249153Sumsquaredresid0.030461Schwarzcriterion-3.050006Loglikelihood36.49153Hannan-Quinncriter.-3.210277「statistic3565.099Durbin-Watsonstat1.119350Prob(F-statistic)0.000000将p的值带入查分模型如下输入:EquationEstimationSpecificationOptions|Equationspeci£icationDependentvariablefoilowedbylistofregressorsandPDLterms;ORexplicit-eq.uationlik«elog(y)"0.68*log(y1))clog:G«)-0.68*logG«("1))Estim^tioxisettingsththod:~~-~~LeastSquares~ftILSandARMA)▼|Sample:119802000~取消确定取消结果如下:

View|P「ou|ObjEut|P「int|Nam亡|Fr亡皂ze|EstimateFonecest|Stats|R^sidsDependentVariable:LOG(Y}-0.627873TOG(Y(-1))Method:LeastSquaresDate:10/17/12Time:19:30Sample(adjusted):19812000Includedobservations:20afteradjustmentsVariableCoefficientStd.Errort-StatisticProt>.C0.4255690.1301073.2709200.0042LOGOO-O.627873*LOG(X(-1))0.9017320.03810223.666440.0000R-squared0.968864Meandependentvar3.480985AdjustedR-squared0.967134S.D.dependentvar0.397883S.E.ofregression0.072132Akaikeinfocriterion-2.325987Sumsquaredresid0.093655Schwarzcriterion-2.226414Loglikelihood25.25987Hannan・Quinncriter.-2.306550F-statistic560.1004Durbin-Watsonstat1.393211Prob(F-statistic)0.000000d.w含糊可以用Im检验其是否任然具有相关性。用b(贝塔)0/(l・p)p二之前的0.6278或者是第一次的:Ld・w/2最小二乘法输入如下EquationEstimationEquationEstimationSpecificationOptions]EquationspecificationDependentvariaLLefollowedbylistofregressorsandPDLteririE^ORanexpliciteq^uationlikelog(y)c“g仗)h(1)3Estimationsettings3Method:|lS~LeastSquares(NLSandAEWA)1980200019802000确走结呆如卜

View|P「x|°bj&:t|Print|Name:|Fr•亡亡ze:|EstimateForecast|Stats|R亡sidsDependentVariable:LOG(Y)Method:LeastSquaresDate:10/17/12Time:19:38.Sample(adjusted):19812000Incl

温馨提示

  • 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
  • 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
  • 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
  • 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
  • 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
  • 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
  • 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。

评论

0/150

提交评论