Eviews操作習題練習要點

1、1 .中國居民人均消費模型從總體上考察中國居民收入與消費支出的關(guān)系。表2.1給出了 1990年不變價格測算的中國人均國內(nèi)生產(chǎn)總值（GDPP）與以居民消費價格指數(shù)（1990年為100）所見的人均居民消費支出（CONSP）兩組數(shù)據(jù)。表2.1中國居民人均消費支出與人均GDP（單位：元/年份CONSPGDPP年份CONSPGDPP1978395.8000675.10001990797.10001602.3001979437.0000716.90001991861.40001727.2001980464.1000763.70001992966.60001949.8001981501.9000792.40

2、0019931048.6002187.9001982533.5000851.100019941108.7002436.1001983572.8000931.400019951213.1002663.7001984635.60001059.20019961322.8002889.1001985716.00001185.20019971380.9003111.9001986746.50001269.60019981460.6003323.1001987788.30001393.60019991564.4003529.3001988836.40001527.00020001690.8003789.7

3、001989779.70001565.900Eviews 操作:FilenewworkfileQuick-empty group錄入數(shù)據(jù)：(1)畫散點圖Quick-group點擊ok進入BoxFlotObj ectTemplateTyptFrame Axi e/ScaleLegend Line/SymbolFill AreaGraph Opt ionsDetails ：ffiaEic 審aphSpeci fiGraph typeGeneralGraphFit lins：Lin & SymbolBarSpikeAreaArei BandMixed th LinewDat Plot Error E

4、ar,HLgJr-LlW . .(Qpg口-以05 4 )ScatterXY LineXi AreaPieBi striluti onQuantile - QuantileBoxplotUndo Edi is取消選擇scatter,點擊確定口回國M Graph: UKTITLED lorkfile: 3:UntitledView I Proc I Object I PrintflddText| Lins/Shade Remove | Template Options ZoomdQ.ncl口 o o o o 5 & a4008001.2001.6002,000CONSP2 2) OLS估計Qui

5、ckestimate equationEqual ion Estisa.t ionXSpecifi cat!on Opti。皿3Equation specification.Dependent variable followeii by list regr*ssorE and FDL term% OR axi explicit equation likegdpp c conspEstination settingsWethod： 二Least Squares(MS and AEBA). |Sample 111978 2000確定 取消 I得出輸出結(jié)果如圖:Equation: UBTITLED

6、forkfile: 3:Untitled UT 叵Y淚向| Prcic Objeqt| Print Name Freeze Estimate Forecast 匕| R+Jd與Dependent Variable: GDPPMethod: Least SquaresDate: 11/14/12 Time: 18:36Sample： 1978 2000Included observations: 23VariableCoefficientStd Error t-StatisticProbC-503.699547.05580-10.704300 0000CONSP2.5705B6D04807153

7、.474710.0000R-squared0.992710Men dependent var1623.530Adjusted R-squared0.992363S D dependent var902.0372S.E. of regression日 5,82266Akaike info criterion11,82536Sum squared resid1546761Schwarz criterion11,92412Log likelihood-133.9919Hannan-Quinn criter11.95022F-statlstlc2859544Durbin-Watson stat0 54

8、3407Prob(F-statistic)0.000000由輸出結(jié)果可以看出，對應(yīng)的回歸表達式為:gdpp = -503.6995 2.5706 consp(-10.704) (53.47)R2 =0.9927F=2859.544DW=0.5434(3)在consp=2000下模型的樣本外預(yù)測方法首先修改工作文件范圍View Object I Print Save Details+Z-5how Fetch 5tore DI:已 Genr SampleR3Sa回0S0Set Sample. a .Structur e/Resi z* Curreiit 一，顏pend to Current F私售

9、號.Contract Current FageB .Reshape Current PageCopy/Extract from Current FageSort Currant Page_ . .Load fforkfile r age a . Save Current Pae .,. Renam Current Page 一 Delete Current PageImportExportDisplay Filter:將工作文件范圍從1978-2000改為1978-2001WorkTile structareWorkfile structure typeDated - regular freq

10、uency 二確定后將工作文件的范圍改為包括 24個觀測值，然后修改樣本范圍Viewssg畫討0*出 用nt 11 Saw | 口心ik+卜| 弓 ho | F白 tch | Sbors | Delete | Genr | SampleSet Sample.S true tur Re sire Curr tut Page. P, 棲p史nd to Current Page.Contract Current Page.Reshape Current FgeCopy/Extract from Current Page Sort Current Page .Load fforkfile Page .

11、 Save Current Page ,. Rftname Current Page .一* Delete Current PageImportExportUntitled / New Page /將范圍樣本改為1978 2001Display Filter: *打開consp的數(shù)據(jù)文件，利用Edit+/-給consp的2001年觀測值賦值 為 2000。I ForecastEcuation: UNTITLEDSeries: GDPPSeries namesForecast name: |gdppfS.E. (optional): |gdppfseGARCH(optional):MethodS

12、tatic forecast(no dynamics in equation) Structural (ignore ARMA)3 Coef uncertainty in S.E. calcOutput3 Forecast graph Q Forecast evaluationv Insert actuals for out-of-sample observationsCancel由上面兩表可以知道：當consp=2000時，gdpp的預(yù)測值為4637.473, 分布標準差為102.2489。6.表2.6給出了 1988年9個工業(yè)國的名義利率（Y）與通貨膨脹率（X）的數(shù)據(jù) （1）以利率為縱軸，

13、以通過膨脹率為橫軸作圖；（2）用OLS法進行回歸分析; （3）如果實際利率不變，則名義利率與通貨膨脹率的關(guān)系如何。表2.6國家Y/%X/%澳大利亞11.97.7加拿大9.44法國7.53.1德國41.611.34.8國家Y/%X/%墨西哥66.351瑞典2.22英國10.36.8美國7.64.4Eviews 操作：建立新的文件:輸入數(shù)據(jù)數(shù)量9:DaU rangeObservations： poorkTile structareWorkfile structure typeUnstructured / Undated二錄入數(shù)據(jù):(1)畫出散點圖:籍 EViewsFile Edit Object

14、工i 電 w Fr&cQui ckOr* i Mis Wi ndow HelpGtaph Empty Group di t Series)Seri Stati stiesForkfile:View Proc Object PrintGro嚨 Statiwti Estimate Equation, 一一 Estimate VAR.eese| DefaultobsYX4111,900007.00000020.7000004.0000007 片nnnnn飛1nnnnnxRange: 19-9 obs | Sample; 19-9 改s國 alphaOlGraph OptionsBoxPlotType

15、Frame | Axis/ScaleObjectLegendTemplateGraph typeGeneralBasic graph |SpecifiLine & SymbolBarSpikeAreaArea BandMixed with LinesDot PlotError BarHighLow (Open-Close)ScatterXY LineXY AreaPieDi stributi onQuantile - QuantileBoxplotLine/SymbolFillUndo Edits取消(2) OLS法進行回歸分析set i tx Proc0t i ons 世indo* Kelp

16、Qui ck Graph:View | Proc |。1Sample.Genrate Series.砧。蝴Graph , P .Empty Group (Edit Seri es)Series StatisticsGrolls Stati stiesEstimate Equation.Estimate VUl.BOEquation EatiationSp*cifi cati on Optj 仙工Equation specification。士p 電 nd 虺 nt vari aI1& fallowed by li st of r egrssor z and ?DLOK an xplicit e

17、quation likey c IFEwtimation settings同ethod：fl- Last Squares (NLS and AENA)Sampled |1 9確定 取消 |Equation: OTITITLED forkfile: IJBTITLED: :- _ - 二 |F1|XVieM | Prem 10bjeizl: | Print | Mam已 | FfEize | EsHm百恒 | Furec石貪 | 專tab | REid專Dependent Variable: YMethod: Least SquaresDate: 11/14/12 Time: 20:41Samp

18、le: 1 SIncluded observations: 9VariableCoefficientStd Error t-StatisticProbC1.55111261.1431671.3508880 2170X1 2731030 06515219 541980.0000R-squared0.982000Man dependent var13.53333Adjusted R-squared0.979429S D dependent var2C18062BE of regression2.694455Akaike info errterion5156601Sum squared resid5 a64509Schwaiz criterion5.20042BLo