求教pesaran CCE估计

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

对于面板数据的横截面相关现在主要有两种处理办法，一个是因子分析，另一个是利用空间相关。pesaran 2006年提出了CCE估计方法，就是使用因子分析来处理截面相关，附件里面是他给出的用于估计的GAUSS程序代码，一调试就出现index of range，看那位大侠能帮忙处理一下。

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏2 回帖

关键词：Pesaran Sara RAN ESA SAR 横截面 程序 空间

PESARAN, M.H., (2006), ESTIMATION and INFERENCE IN LARGE HETEROGENEOUS PANELS WITH A MULTIFACTOR
  ERROR STRUCTURE, ECONOMETRICA, VOL74, NO.4, PP.967-1012.
**********************************************************************************************************
++++++++++++++++ <Preliminary Information on Panel Data> ++ ++++++++++++++++++++++++++++++++++++++++++++

Min(year  Max(Year   Min(Ti)   Max(Ti)         N   mean Ti
     1970      1979         9        10         3      9.33

++++++++++++++++ <OLS Mean Group Estimates> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

               y            coeff           se(NP)            t(NP)
              x1         0.908876         0.663846         1.369106
              x2         0.374051         0.072670         5.147241
              d1         1.006693         0.132861         7.577001
            inpt         2.449062         1.227480         1.995194

   Pesaran's (2004) CD test statistic (residuals): -1.59442
     (Pesaran, M.H., (2004), General Diagnostic Tests for Cross Section Dependence in Panels, CESifo
       Working Papers No.1233)
   * OLS Mean Group Estimator is simple cross section average of OLS estimator for each cross section unit.
   * se(NP) is the standard error based on non-parametric variance estimator of eq(58) in Pesaran (2006).
   *  t(.) is associated t-ratio.

++++++++++++++++ <Fixed Effects Estimates> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

        y     coeff    se(NP)     t(NP)    se(NW)     t(NW)
       x1  0.808984  0.628055  1.288076  0.486449  1.663040
       x2  0.217298  0.181566  1.196801  0.435513  0.498947

   * se(NP) is the standard error based on non-parametric variance estimator of eq(69) in Pesaran (2006).
   * se(NW) is the standard error based on Newey-West type variance estimator of eq(74) in Pesaran (2006).
   *  t(.) is associated t-ratio.

   Cross Section Averages of Estimated Heterogeneous Slops in FE Estimation

       d1  1.026741
     inpt  2.445495

++++++++++++++++ <Pesran's (2006) CCE Mean Group Estimates> ++++++++++++++++++++++++++++++++++++++++++++

               y            coeff           se(NP)            t(NP)
              x1         1.599212         0.639212         2.501846
              x2         -2.01574         0.391813         -5.14464

   * CCE Mean Group Estimator is defined by eq(53).
   * se(NP) is the standard error based on non-parametric variance estimator of eq(58) in Pesaran (2006).
   *  t(.) is associated t-ratio.

   Cross Section Averages of Estimated Heterogeneous Slops in CCE Mean Group Estimation

              d1         -1.97231
            inpt         -4.79778
            ybar         3.001323
           x1bar         -1.84820
           x2bar         -1.49441

++++++++++++++++ <Pesaran's (2006) CCE Pooled Estimates> +++++++++++++++++++++++++++++++++++++++++++++++

        y     coeff    se(NP)     t(NP)    se(NW)     t(NW)
       x1  1.253333  0.896996  1.397255  0.479063  2.616214
       x2  -1.26680  0.720385  -1.75850  0.521113  -2.43095

   * CCE Pooled Estimator is defined by eq(65).
   * se(NP) is the standard error based on non-parametric variance estimator of eq(69) in Pesaran (2006).
   * se(NW) is the standard error based on Newey-West type variance estimator of eq(74) in Pesaran (2006).
   *  t(.) is associated t-ratio.

   Cross Section Averages of Estimated Heterogeneous Slops in CCE Pooled Estimation

       d1  -1.13908
     inpt  -3.16088
     ybar  2.148750
    x1bar  -0.94997
    x2bar  -0.33588

--------------------------------------------------------------
N used for Cross Section Average (out of        3 )

     1970         2
     1971         3
     1972         3
     1973         3
     1974         3
     1975         3
     1976         3
     1977         3
     1978         3
     1979         2
--------------------------------------------------------------
Reports on CCE Estimates of All Cross Section Units


       ID        x1    se(NW)        x2    se(NW)        d1    se(NW)      inpt    se(NW)      ybar    se(NW)     x1bar    se(NW)     x2bar    se(NW)        Ti      From        To
        1     2.000     0.312    -1.624     0.551     0.756     0.094    -1.313     0.346     0.514     0.099    -1.919     1.036     3.818     0.297         9      1970      1978
        3     0.347     0.207    -2.016     0.000     0.830     0.307     3.150     0.339     0.155     0.283     0.869     0.809    -2.477     0.672        10      1970      1979
        5     2.450     0.049    -2.408     0.056    -7.503     0.223   -16.231     0.661     8.336     0.223    -4.495     0.156    -5.824     0.227         9      1971      1979

   * CCE Estimator for a cross section unit is defined by eq(26).
   * se(NW) is the standard error based on Newey-West type variance estimator of eq(50) in Pesaran (2006).

路径设了没？你用GAUSS7.0

Home     People   Research     Intranet    About   Seminars   News    Computing    Prospectus Library

You are in:  Home &raquo; Staff

Staff Academic Staff University Teaching Officers Research Officers College Teaching Officers Emeritus Professors Visitors/Honorary PhD Students Admin & Support Where are they now Ex-Academic Staff Retired Academic Staff Ex-PhD Students

[td]University Teaching Officers M. Hashem Pesaran FBA Professor of Economics College: Fellow of Trinity College Email: mhp1@cam.ac.uk Tel: +44-(0) 1223-335216 (Faculty Office) +44-(0) 1223 338403 (Trinity Office) Room: 16 Office Hours: Wednesdays 2-4:30 Subject to making an appointment with Sharon Swann (ss243@cam.ac.uk) Interests: Econometric Analysis of Heterogeneous Panels with Unobserved Common Effects. Panel unit root tests. Analysis of Panel Vector Autoregressive Models(PVAR). Long-run Structural Macroeconometric Modelling. Global Vector Autoregressive Modelling (GVAR). Economic and Financial Forecasting in the Presence of Structural Breaks. Financial Econometrics - credit risk analysis and portfolio optimization. Econometric Analysis of Non-nested Models. Empirics of Growth. Research on EconPapers ** Current CV Teaching Personal Pages My Publications Biographical Editorial Research Projects Microfit 5 New Audio, Video and Podcasts New Journal of Applied Econometrics Journal of Applied Econometrics Fall 2010 Newsletter New Call for Papers Iran Economy - IIEA Conference Sept 2011 New The GVAR Toolbox 1.1 New A listing of my work available on EconPapers. You may find more specific links to my work in my Personal Pages box

[/td][/tr]
[/table]

©2011 Faculty of Economics, University of Cambridge, Sidgwick Avenue, Cambridge, CB3 9DD	Information provided by webmaster@econ.cam.ac.uk.

PesaranCVDec2010.rar (143.5 KB) 本附件包括：

• PesaranCVDec2010.pdf

2011-10-21 22:23:07 上传

Hashem Pesaran.doc (31 KB)

2011-10-21 22:22:50 上传

LN-SR052010-EN.rar (568.86 KB) 本附件包括：

• LN-SR052010-EN.pdf

2011-10-21 22:22:27 上传

xuehe 发表于 2011-10-21 21:57
路径设了没？你用GAUSS7.0

谢谢你的回复。路径按他给的PDF文件中的说明设置好了，就是一运行，就会出现index of range，一直不知道怎么解决这个问题

路径如下
output file = D:\\gausscd\\CCEgauss7_Aug\\CCEgauss7_out.txt RESET;

/************************************************************/
/** load cross section id, years, y_it, x_it (first sheet) **/
/************************************************************/
    /* Specify the rage of first raw, variables names */
        vname=xlsreadm("D:\\gausscd\\CCEgauss7_Aug\\CCEtestdata.xls","a1:e1",1,"");

    /* Specify the range of the cross section varying regresssors, second raw onwards */
        data=xlsreadm("D:\\gausscd\\CCEgauss7_Aug\\CCEtestdata.xls","a2:e29",1,"");

xuehe 发表于 2011-10-22 17:09
路径如下
output file = D:\\gausscd\\CCEgauss7_Aug\\CCEgauss7_out.txt RESET;

请问你这样改过路径滞后能正常运行吗？路径我也按照要求修改过了
output file = C:\\gauss\\CCEgauss6_22Aug08\\CCEgauss6_out.txt RESET
/* Specify the rage of first raw, variables names */
        vname=xlsreadm("C:\\gauss\\CCEgauss6_22Aug08\\CCEtestdata.xls","a1:e1",1,"");

    /* Specify the range of the cross section varying regresssors, second raw onwards */
        data=xlsreadm("C:\\gauss\\CCEgauss6_22Aug08\\CCEtestdata.xls","a2:e29",1,"");
但还是不行，

可能存在别的问题，就不知道了？

output file = D:\\gausscd\\CCEgauss7_Aug\\CCEgauss7_out.txt RESET;

少；？