ywh19860616 发表于 2012-1-2 12:02 
epoh老师,您上面写的XS1 是表示什么?
可能我没有想到那么复杂,我的想法是现在要同时检验XS(-1)和XS(- ...
哈哈!老兄原来你才是罪魁祸首.
42楼你写错了
FRML eq. ys.=a.+b1.*ys(-1)+b2.*ys(-2)+c1.*xs.(-1)+c2.*xs(-2)
应该是
FRML eq. ys.=a.+b1.*ys.(-1)+b2.*ys.(-2)+c1.*xs.(-1)+c2.*xs.(-2);
这才折腾了半天.呵呵
现在程序可以运行了,你试试.
请注意短信息
不过如果要做bootstrap
PROC RESIDNUL beta eps;
还需要修改
因为原来是
FRML eq0. ys.=a0.+b01.*ys.(-1);
PARAM a0. b01.;
*******************
FRML constr1. c1.;
FRML constr2 c2.;
ANALYZ constr1.;
ANALYZ constr2.;
和
FRML constr1. c1.;
FRML constr2 c2.;
ANALYZ constr1. constr2.;
是不一样的
底下例子(MSGM_modified)可以说明:
MSGM_modified.tsp,HedPrice1.dat,Team.dat
frml c1 d1;
frml c2 d2;
frml c3 beta;
? Benchmark estimation with Wald test of separability
LSQ(MAXIT=500,TOL=0.0001) E1 E2 E3 E4;
set loglu=@logl;
analyz c1 c2 c3;
Results of Parameter Analysis
=============================
Standard
Parameter Estimate Error t-statistic P-value
C1 13.4943 18.7142 .721072 [.471]
C2 292.864 214.771 1.36361 [.173]
C3 18.7071 5.50715 3.39687 [.001]
Wald Test for the Hypothesis that the given set of Parameters are jointly zero:
CHISQ(3) = 19.941252 ; P-value = 0.00017
*******************
LSQ(MAXIT=500,TOL=0.0001) E1 E2 E3 E4;
set loglu=@logl;
analyz c1 ;
analyz c2 ;
analyz c3 ;
Results of Parameter Analysis
=============================
Standard
Parameter Estimate Error t-statistic P-value
C1 13.4940 18.7140 .721067 [.471]
Wald Test for the Hypothesis that the given set of Parameters are jointly zero:
CHISQ(1) = 0.51993751 ; P-value = 0.47087
Results of Parameter Analysi
=============================
Standard
Parameter Estimate Error t-statistic P-value
C2 292.867 214.769 1.36363 [.173]
Wald Test for the Hypothesis that the given set of Parameters are jointly zero:
CHISQ(1) = 1.8594975 ; P-value = 0.17268
Results of Parameter Analysis
=============================
Standard
Parameter Estimate Error t-statistic P-value
C3 18.7071 5.50717 3.39687 [.001]
Wald Test for the Hypothesis that the given set of Parameters are jointly zero:
CHISQ(1) = 11.538697 ; P-value = 0.00068