eviews期中作业报告(要点基本统计和OLS稳健方差)

Eviews期中实验报告一、实验任务上机内容：基本统计和OLS、稳健方差要点∙描述性统计∙简单假设检验∙季节调整：移动平均法∙多元统计分析：齐性检验；主成分分析∙方程对象：方程设定，估计结果，系数，成员函数∙方差稳健估计：HC和HAC∙哑变量和交互项二、实验内容1.简单统计分析与描述性统计数据查看：m1.sheetm1StatsM1Mean378.1643Median274.0275Maximum1089.475Minimum129.8910Std.Dev.265.4934Skewness1.012404Kurtosis2.765155Jarque-Bera27.70001Probability0.000001Sum60506.30SumSq.Dev.11207393Observations160从而显示出了m1的各种统计信息，包括均值，中位点，最大值最小值，以及二到四阶矩观测数目和正态分布检验等的统计概要。M1.line则显示出了这一组数据的图形描述freeze(gk)G(1).distplotkernel'G(1).kdensity'V6-G(1).distplotkernel'画出对象g中第一个变量的直方图，命名为ghfreeze(gh)G(1).hist'histlog(m1)02004006008001,0001,20019551960196519701975198019851990M1.0.1.2.3.4.5.64.04.55.05.56.06.57.07.58.0DensityLOG(M1)'将g中的所有变量画入一张线性图中，改图命名为gfagraphgfa.lineG'将g中所有的变量分别画出一张图并合并入gfg这一个对象中graphgfg.line(m)G'multiplegraphs'生成列表tbgsi，该列表中包含了g中所有变量单独的统计值freeze(tbGsi)G.stats(i)'individualsamples0510152025304.85.05.25.45.65.86.06.26.46.66.87.07.2Series:LOG(M1)Sample1952Q11996Q4Observations180Mean5.811220Median5.698431Maximum7.106131Minimum4.840535Std.Dev.0.754650Skewness0.323130Kurtosis1.694566Jarque-Bera15.91357Probability0.000350-40481216556065707580859095LOG(M1)LOG(GDP)RSDLOG(PR)Date:11/17/12Time:20:04Sample:1952Q11996Q4LOG(M1)LOG(GDP)RSDLOG(PR)Mean5.811225.9915055.4129280.009645Median5.6984315.9250095.05750.008295Maximum7.1061317.57467415.087330.030557Minimum4.8405354.4759150.814333-0.00097Std.Dev.0.754651.0025332.9089390.006206Skewness0.323130.0623610.9867820.909753Kurtosis1.6945661.5629714.0498833.466402Jarque-Bera15.9135715.6045737.4790726.31399Probability0.000350.00040900.000002Sum1046.021078.471974.3271.72653SumSq.Dev.101.9398179.90791514.6850.006855Observations180180180179回归模型估计smpl1952Q11992Q4equationeq1.lslog(m1)clog(gdp)rsdlog(pr)DependentVariable:LOG(M1)Method:LeastSquaresDate:12/07/12Time:15:41Sample(adjusted):1952Q21992Q4Includedobservations:163afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.C1.3123830.03219940.758500.0000LOG(GDP)0.7720350.006537118.10920.0000RS-0.0206860.002516-8.2211960.0000DLOG(PR)-2.5722040.942556-2.7289670.0071R-squared0.993274Meandependentvar5.692279AdjustedR-squared0.993147S.D.dependentvar0.670253S.E.ofregression0.055485Akaikeinfocriterion-2.921176Sumsquaredresid0.489494Schwarzcriterion-2.845256Loglikelihood242.0759Hannan-Quinncriter.-2.890354F-statistic7826.904Durbin-Watsonstat0.140967Prob(F-statistic)0.000000这里我们使用smpl来设定估计样本的观测区间，而equation语句则创建了方程对象eq1并进行最小二乘估计，其中的c代表回归方程的常数项，得到模型估计的结果如上图所示。freeze(gfeq1r)eq1.resids这里我们产生了残差项的图使用表格来查看残差，命令为：eq.Resids（t）另外我们可以查看回归结果的文本表示：eq1.representationsEstimationCommand:=========================LSLOG(M1)CLOG(GDP)RSDLOG(PR)EstimationEquation:=========================LOG(M1)=C(1)+C(2)*LOG(GDP)+C(3)*RS+C(4)*DLOG(PR)SubstitutedCoefficients:=========================LOG(M1)=1.31238347449+0.772034899215*LOG(GDP)-0.0206860343222*RS-2.57220371427*DLOG(PR)接下来我们来对这一结果进行假设检验。进行wald检验，检验b4=2是否成立，命令为：Eq1.waldc(4)=2WaldTest:Equation:EQ1TestStatisticValuedfProbability-.15-.10-.05.00.05.10.154.55.05.56.06.57.07.519551960196519701975198019851990ResidualActualFittedF-statistic23.53081(1,159)0.0000Chi-square23.5308110.0000NullHypothesisSummary:NormalizedRestriction(=0)ValueStd.Err.-2+C(4)-4.5722040.942556Restrictionsarelinearincoefficients.从而我们看到这一假设被拒绝了，下面我们进行序列相关检验，命令如下：eq1.auto(1)Breusch-GodfreySerialCorrelationLMTest:F-statistic813.0060Prob.F(1,158)0.0000Obs*R-squared136.4770Prob.Chi-Square(1)0.0000TestEquation:DependentVariable:RESIDMethod:LeastSquaresDate:12/07/12Time:15:49Sample:1952Q21992Q4Includedobservations:163Presamplemissingvaluelaggedresidualssettozero.VariableCoefficientStd.Errort-StatisticProb.C-0.0063550.013031-0.4876830.6265LOG(GDP)0.0009970.0026450.3769290.7067RS-0.0005670.001018-0.5567480.5785DLOG(PR)0.4041430.3816761.0588640.2913RESID(-1)0.9203060.03227628.513260.0000R-squared0.837282Meandependentvar4.92E-16AdjustedR-squared0.833163S.D.dependentvar0.054969S.E.ofregression0.022452Akaikeinfocriterion-4.724644Sumsquaredresid0.079649Schwarzcriterion-4.629744Loglikelihood390.0585Hannan-Quinncriter.-4.686116F-statistic203.2515Durbin-Watsonstat1.770965Prob(F-statistic)0.000000接下来我们要进行模型的修改，调整季节性，采用移动平均法equationeq2.lslog(m1)clog(gdp)rsdlog(pr)log(m1(-1))log(gdp(-1))rs(-1)dlog(pr(-1))DependentVariable:LOG(M1)Method:LeastSquaresDate:12/07/12Time:15:52Sample(adjusted):1952Q31992Q4Includedobservations:162afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.C0.0712970.0282482.5239490.0126LOG(GDP)0.3203380.1181862.7104530.0075RS-0.0052220.001469-3.5548010.0005DLOG(PR)0.0386150.3416190.1130360.9101LOG(M1(-1))0.9266400.02031945.603750.0000LOG(GDP(-1))-0.2573640.123264-2.0879100.0385RS(-1)0.0026040.0015741.6544290.1001DLOG(PR(-1))-0.0716500.347403-0.2062460.8369R-squared0.999604Meandependentvar5.697490AdjustedR-squared0.999586S.D.dependentvar0.669011S.E.ofregression0.013611Akaikeinfocriterion-5.707729Sumsquaredresid0.028531Schwarzcriterion-5.555255Loglikelihood470.3261Hannan-Quinncriter.-5.645823F-statistic55543.30Durbin-Watsonstat2.393764Prob(F-statistic)0.000000equationeq2.lslog(m1)clog(gdp)rsdlo