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
1、 面板数据模型的Views操作 (EViews Operation of Panel data Model)Pooled Time Series, Cross-Section DataData ofen contain information on a relatively small number of cross-sectional unit observed over time. For example,you may have time series data on GDP for a number of European nations.Or perhaps you have sta
2、te level data on unemployment observed over time.We term such data pooled time series,cross-section data.EViews provides a number of specialized tools to help you work with pooled data. EViews will help you manage your data,perform operations in either the the time series or the cross-section dimens
3、ion,and apply estimation methods that accpunt for the pooled structure of your data.EViews Object that manages time series/cross-section data is called a pool. The experiment will describe how to set up your data to work with,and how to define and work with objects.【实验目的】掌握面板数据模型基本内容的软件操作【实验内容】面板数据模
4、型的实验内容:建立面板数据工作文件;对面板数据的处理;面板数据模型的参数估计一、从Excel数据导入建立一个工作文件1双击Views标识,打开Views主窗口;从Views主菜单中点击File键,选择openforeign data as Workfile单击左键弹出一个窗口找到Excel数据表点击打开点击下一步点击完成。2得到一个工作文件,并弹出一个工作文件表格点击文件表格name点击OK键。3对工作文件保存、命名,点击save(保存)点击OK键。二、建立面板数据库并处理数据 向EViews6.0中输入截面数据名称的时候,应先建立一个合并数据(Pool)对象。 The most direst
5、 way of creating a pool object is to select Object.new Object/pool.选择EViews6.0主菜单ObjectNew ObjectPool 在Pool中输入_BJ_TJ_HB_LN_SHH_JS_ZHJ_FJ_SHD_GD_HN在Pool窗口点击name,保存。在Pool窗口点击sheet,打开一个窗口,输入GDP?,RENKOU?,GSH?,GZH?。就得到一个东部地区GDP,RENKOU,GSH,GZH的Poolsheet(面板数据表)。在Pool窗口点击define,回到Pool的标示窗口;点击Pool的标示窗口sheet,
6、打开一个窗口,输入GDP?,RENKOU?,GSH?,GZH?。得到各个个体GDP,RENKOU,GSH,GZH的Poolsheet(面板数据表)。Pool序列的序列名使用的是基本名和“?”占位符。例如,GDP?代表:GDP_BJ北京GDPGDP_TJ天津GDPGDP_HB河北GDPGDP_LN辽宁GDPGDP_SHH上海GDPGDP_JS江苏GDPGDP_ZHJ浙江GDPGDP_FJ福建GDPGDP_SHD山东GDPGDP_GD广东GDPGDP_HN海南GDP还可以通过Pool窗口中的PoolGenerate,通过公式可以生成以面板数据为基础的新数据。例如,RJGDP?=GDP?/RENKO
7、U?RJGDP_BJ北京人均GDPRJGDP_TJ天津人均GDPRJGDP_HB河北人均GDPRJGDP_LN辽宁人均GDPRJGDP_SHH上海人均GDPRJGDP_JS江苏人均GDPRJGDP_ZHJ浙江人均GDPRJGDP_FJ福建人均GDPRJGDP_SHD山东人均GDPRJGDP_GD广东人均GDPRJGDP_HN海南人均GDP面板数据的处理1计算各个个体GDP占全体GDP的比重步骤(1)构造sum在普通group里构造sum,即构造一序列sum,sum全部变为0,如果为NA的话,则处于不能计算的状态,要注意把sum下面的NA全部变为0。(2)打开面板数据Pool在PoolGener
8、ate输入:sum=sum+GDP?(把各个个体的横截面GDP加总)(3)把sum放入Pool。即打开Pool,点击define,显示编辑窗口,在这个窗口点击sheet,输入sum,点击OK。(4)计算比重,打开面板数据Pool,在PoolGenerate输入BZHGDP?=GDP?/sum2计算任意两点之间的平均增长速度比方说,计算1978-2008年的年均增长速度打开面板数据Pool,在PoolGenerate输入:suduGDP?=(GDP?/GDP?(-30)(1/30)-1点击OK键三、面板数据模型的参数估计利用合并数据库(Pool)进行参数估计点击合并数据库(Pool)工具栏中的E
9、stimate,出现对话框。(如果要把计算机画面全屏复制下来,操作Shift+Print,单击鼠标左键粘贴)(一)混合模型(Pooled model)的估计方法 yit = a + Xit 'b +eit, i = 1, 2, , N; t = 1, 2, , T打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? 在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)cross-cection对应的方框,为none。period对应的方框,为none。weights对
10、应的方框,为no weights 。其他的不填。点击OK键Dependent Variable: GSH?Method: Pooled Least SquaresDate: 12/20/09 Time: 22:51Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341VariableCoefficientStd. Errort-StatisticProb. C15.8540911.792321.3444420.1
11、797GDP?0.0805260.00169847.426670.0000R-squared0.869025 Mean dependent var326.2877Adjusted R-squared0.868639 S.D. dependent var499.7634S.E. of regression181.1330 Akaike info criterion13.24219Sum squared resid11122307
12、0; Schwarz criterion13.26466Log likelihood-2255.793 Hannan-Quinn criter.13.25114F-statistic2249.289 Durbin-Watson stat0.084506Prob(F-statistic)0.000000在左上方Dependent Variable(被解释变量)选择框内填GSH? 在Common coefficients(相同系数) 选择框内填C GDP? TZ?cross-cection对应的方
13、框,为none。period对应的方框,为none。weights对应的方框,为no weights 。其他的不填。点击OK键Dependent Variable: GSH?Method: Pooled Least SquaresDate: 12/20/09 Time: 22:53Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341VariableCoefficientStd. Errort-StatisticProb.
14、 C11.6724511.772920.9914660.3222GDP?0.0987090.00674514.634450.0000TZ?-0.0425020.015269-2.7835320.0057R-squared0.871961 Mean dependent var326.2877Adjusted R-squared0.871203 S.D. dependent var499.7634S.E. of regression179.3567 A
15、kaike info criterion13.22539Sum squared resid10873061 Schwarz criterion13.25910Log likelihood-2251.929 Hannan-Quinn criter.13.23882F-statistic1150.905 Durbin-Watson stat0.089560Prob(F-statistic)0.000000因为Pool只是一个虚拟的窗口,在此窗口下保留回归的结果,
16、若要回到数据表状态,双击这个窗口,点击sheet即可。(二)个体固定效应回归模型的估计方法(entity fixed effects model)固定效应变截距模型(fixed effects regression model)。固定效应模型分为3 种类型,即个体固定效应变截距模型、时点固定效应变截距模型和个体时点双固定效应变截距模型 yit = ai + Xit 'b +eit, i = 1, 2, , N; t = 1, 2, , T1个体固定效应模型也可以表示为 yit = a1 D1 + a2 D2 + +aN DN + Xit 'b +eit, t = 1, 2, ,
17、 Ty1t = a1 + X1t 'b +e1t,y2t = a2 + X2t 'b +e2 t,yN t = aN + XN t 'b+e N t,打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimation Method : 在cross section 填fixed 在period填 no 在weights填 no weights在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependent
18、Variable: GSH?Method: Pooled Least SquaresDate: 12/20/09 Time: 23:02Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341VariableCoefficientStd. Errort-StatisticProb. C7.09525310.030310.7073810.4798GDP?0.0827980.00152154.423080.0000Fi
19、xed Effects (Cross)_BJ-C136.7429_TJ-C15.92551_HB-C-103.9208_LN-C14.87019_SHH-C230.4823_JS-C-87.77667_ZHJ-C-47.23187_FJ-C-36.83428_SHD-C-157.7241_GD-C41.98773_HN-C-6.521002Effects SpecificationCross-section fixed (dummy variables)R-squared0.912530 Mean dependent var326.2877Adju
20、sted R-squared0.909606 S.D. dependent var499.7634S.E. of regression150.2571 Akaike info criterion12.89713Sum squared resid7427893. Schwarz criterion13.03197Log likelihood-2186.960 Hannan-Quinn criter.12.95085
21、F-statistic312.0274 Durbin-Watson stat0.123789Prob(F-statistic)0.0000002时点固定效应回归模型的估计方法(time fixed effects model) yit = gt + Xit 'b +eit, i = 1, 2, , N时点固定效应变截距模型也可以加入虚拟变量表示为yit =g0 + g1 W1 + g2 W2 + +g T WT + Xit 'b +eit, i = 1, 2, , N; t = 1, 2, , Tyi1 = (g0 + g1) +
22、X1t 'b+ ei1,yi2 = (g0 + g2) + X2t 'b + ei2,yiT = (g0 + gT) + XN t 'b + eiT,打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimation Method : 在cross section 填no;在period填 fixed ;在weights填 no weights。在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependen
23、t Variable: GSH?Method: Pooled Least SquaresDate: 12/20/09 Time: 23:08Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341VariableCoefficientStd. Errort-StatisticProb. C58.4448213.754934.2490090.0000GDP?0.0694780.00263726.350330.0000
24、Fixed Effects (Period)1978-C-14.261921979-C-17.754011980-C-19.251371981-C-20.739391982-C-22.811201983-C-25.629951984-C-24.318481985-C-19.576591986-C-17.239321987-C-18.749521988-C-24.429481989-C-20.151211990-C-23.285271991-C-19.482441992-C-28.144441993-C-13.324551994-C-86.629971995-C-108.20351996-C-1
25、03.98951997-C-100.49351998-C-84.521741999-C-67.399542000-C-49.142122001-C-3.9684842002-C7.3817712003-C18.760512004-C23.282102005-C81.530592006-C133.48752007-C288.19462008-C380.8603Effects SpecificationPeriod fixed (dummy variables)R-squared0.893443 Mean dependent var326.2877Ad
26、justed R-squared0.882753 S.D. dependent var499.7634S.E. of regression171.1261 Akaike info criterion13.21182Sum squared resid9048798. Schwarz criterion13.57141Log likelihood-2220.615 Hannan-Quinn criter.13.355
27、09F-statistic83.57579 Durbin-Watson stat0.061115Prob(F-statistic)0.0000003个体时点固定效应变截距回归模型的估计方法(time and entity fixed effects model) yit = a0 +ai +gt + Xit 'b +eit, i = 1, 2, , N; t = 1, 2, , T个体时点固定效应变截距模型还可以表示为yit = a0 +a1 D1+a2 D2 +aN DN + g1W1+ g2W2 +g TWT + Xit 'b
28、+eit,打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimation Method : 在cross section 填fixed;在period填 fixed ;在weights填 no weights。在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependent Variable: GSH?Method: Pooled Least SquaresDate: 12/20/09 Time: 23:10Sample: 1
29、978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341VariableCoefficientStd. Errort-StatisticProb. C52.1924612.851204.0612910.0001GDP?0.0711000.00271326.210880.0000Fixed Effects (Cross)_BJ-C119.5505_TJ-C-13.56755_HB-C-105.8635_LN-C7.986251_SHH
30、-C224.7008_JS-C-55.49023_ZHJ-C-36.70717_FJ-C-52.38527_SHD-C-125.8993_GD-C85.03988_HN-C-47.36439Fixed Effects (Period)1978-C-8.2676351979-C-11.791131980-C-13.324691981-C-14.842481982-C-16.953811983-C-19.818111984-C-18.597211985-C-13.972611986-C-11.715321987-C-13.372701988-C-19.299581989-C-15.17554199
31、0-C-18.443721991-C-14.887681992-C-23.972681993-C-9.8952471994-C-84.152491995-C-106.73421996-C-103.33241997-C-100.55401998-C-85.123131999-C-68.525742000-C-51.199852001-C-6.8997392002-C3.4105752003-C13.121312004-C15.203242005-C70.391602006-C119.42032007-C270.35952008-C358.9452Effects SpecificationCros
32、s-section fixed (dummy variables)Period fixed (dummy variables)R-squared0.933006 Mean dependent var326.2877Adjusted R-squared0.923819 S.D. dependent var499.7634S.E. of regression137.9389 Akaike info criterion12.80639Sum squared res
33、id5689114. Schwarz criterion13.27836Log likelihood-2141.490 Hannan-Quinn criter.12.99443F-statistic101.5629 Durbin-Watson stat0.097206Prob(F-statistic)0.000000(三)个体随机效应变截距回归模型的估计方法 yit = ai + Xit'b +eit, i = 1, 2, , N; t = 1, 2
34、, , T同理也可定义时点随机效应变截距模型和个体时点随机效应变截距模型,但个体随机效应变截距模型最为常用。打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimation Method : 在cross section 填Random;在period填 no ;在weights填 no weights。在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependent Variable: GSH?Method: Pooled E
35、GLS (Cross-section random effects)Date: 12/20/09 Time: 23:18Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341Swamy and Arora estimator of component variancesVariableCoefficientStd. Errort-StatisticProb. C7.71033833.511630.2300790.
36、8182GDP?0.0826390.00151454.592830.0000Random Effects (Cross)_BJ-C128.2089_TJ-C14.57945_HB-C-97.62742_LN-C13.87791_SHH-C216.3951_JS-C-82.02634_ZHJ-C-44.22540_FJ-C-34.79399_SHD-C-147.7269_GD-C39.98641_HN-C-6.647729Effects SpecificationS.D. Rho Cross-section random106.06810.3326Id
37、iosyncratic random150.25710.6674Weighted StatisticsR-squared0.897846 Mean dependent var80.45430Adjusted R-squared0.897545 S.D. dependent var469.4939S.E. of regression150.2783 Sum squared resid7655831.F-statistic2979.534
38、 Durbin-Watson stat0.120273Prob(F-statistic)0.000000Unweighted StatisticsR-squared0.868427 Mean dependent var326.2877Sum squared resid11173094 Durbin-Watson stat0.082411打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimatio
39、n Method : 在cross section 填no;在period填 Random;在weights填 no weights。在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependent Variable: GSH?Method: Pooled EGLS (Period random effects)Date: 12/20/09 Time: 23:20Sample: 1978 2008Included observations: 31Cross-sections included: 11Total p
40、ool (balanced) observations: 341Swamy and Arora estimator of component variancesVariableCoefficientStd. Errort-StatisticProb. C22.2844913.236511.6835620.0932GDP?0.0788580.00179843.848560.0000Random Effects (Period)1978-C6.6224671979-C5.4302141980-C4.8762971981-C4.3375161982-C3.5909561983-
41、C2.5906671984-C2.8462911985-C4.1650671986-C4.7734841987-C4.0071141988-C1.7003051989-C2.7992611990-C1.5304291991-C2.3013601992-C-1.3037331993-C2.1123171994-C-23.464561995-C-32.358281996-C-32.515161997-C-32.727581998-C-28.559271999-C-23.987522000-C-19.810742001-C-6.7897512002-C-5.0580872003-C-4.496019
42、2004-C-7.6077262005-C5.5523802006-C16.918202007-C60.054192008-C82.46991Effects SpecificationS.D. Rho Period random35.764470.0419Idiosyncratic random171.12610.9581Weighted StatisticsR-squared0.841678 Mean dependent var268.1645Adjusted R-squared0.841211
43、0; S.D. dependent var443.5669S.E. of regression176.7541 Sum squared resid10591038F-statistic1802.205 Durbin-Watson stat0.078287Prob(F-statistic)0.000000Unweighted StatisticsR-squared0.868653 Mean dependent var326.2
44、877Sum squared resid11153972 Durbin-Watson stat0.085847打开Pooled Estimation窗口,见下图。在左上方Dependent Variable(被解释变量)选择框内填GSH? Estimation Method : 在cross section 填Random ;在period填 Random;在weights填 no weights。在Common coefficients(相同系数) 选择框内填C GDP? (可以增加TZ?)其他的不填。点击OK键Dependent Va
45、riable: GSH?Method: Pooled EGLS (Two-way random effects)Date: 12/20/09 Time: 23:21Sample: 1978 2008Included observations: 31Cross-sections included: 11Total pool (balanced) observations: 341Swamy and Arora estimator of component variancesVariableCoefficientStd. Errort-StatisticProb. C19.2
46、258435.662160.5391100.5902GDP?0.0796520.00185542.932410.0000Random Effects (Cross)_BJ-C125.3488_TJ-C7.582734_HB-C-99.09180_LN-C12.35144_SHH-C217.1972_JS-C-75.03942_ZHJ-C-42.12580_FJ-C-38.91561_SHD-C-141.5206_GD-C50.82344_HN-C-16.61030Random Effects (Period)1978-C13.095651979-C11.025631980-C10.057971
47、981-C9.1182491982-C7.8166031983-C6.0746051984-C6.4917161985-C8.7396761986-C9.7696701987-C8.4042201988-C4.3479821989-C6.2057311990-C3.9751161991-C5.2403051992-C-1.1089941993-C4.5935301994-C-39.890011995-C-55.543951996-C-56.038141997-C-56.602421998-C-49.543891999-C-41.783422000-C-34.817502001-C-12.544
48、072002-C-9.8355012003-C-9.3215582004-C-15.371372005-C6.5423652006-C25.390002007-C98.937552008-C136.5743Effects SpecificationS.D. Rho Cross-section random106.60660.3486Period random47.027140.0678Idiosyncratic random137.93890.5836Weighted StatisticsR-squared0.844652 Mean dependent var71.55205Adjusted R-squared0.844193
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。
最新文档
- 采购部采购水电制度
- 采购预付货款管理制度
- 采购验收管理制度细则
- 钢筋材料采购制度
- 2025-2026学年 新人教版数学 八年级下册 第一次月考试卷(原卷)
- 数学早读(课件)-2025-2026学年苏教版一年级数学上册
- 2026年农村姐弟建房合同(1篇)
- 专家认证施工方案(3篇)
- 书法练字营销方案(3篇)
- 企业防盗应急预案(3篇)
- 《中医辩证施护》课件
- 幕墙技术标(暗标)
- 管理会计学 第10版 课件 第6章 存货决策
- 高等代数试卷
- 三方协议解约函电子
- 三对三篮球赛记录表
- 电气自动化社会实践报告
- 【关于某公司销售人员招聘情况的调查报告】
- 拉肚子的故事知乎拉黄稀水
- JJF 1083-2002光学倾斜仪校准规范
- GB/T 2504-1989船用铸钢法兰(四进位)
评论
0/150
提交评论