老师 你好 我自己貌似已经解决了上边的问题了 在givewin里面自己输入数据就会出来那些.in7和.bn7了
Ox version 3.40 (Windows) (C) J.A. Doornik, 1994-2004
MSVAR (c) H-M Krolzig, 1996-2005, package version 1.32a, object created on 29-11-2004
---------- EM algorithm converged after 98 iterations ------------
EQ( 1) MSMH(3)-VAR(1) model of (neer,gdp,cpi,wcpi)
Estimation sample: 1996 (2) - 2011 (4)
no. obs. per eq. : 183 in the system : 732
no. parameters : 64 linear system : 30
no. restrictions : 28
no. nuisance p. : 6
log-likelihood : -2156.2373 linear system : -2263.3551
AIC criterion : 24.2649 linear system : 25.0640
HQ criterion : 24.7199 linear system : 25.2773
SC criterion : 25.3873 linear system : 25.5901
LR linearity test: 214.2356 Chi(28) =[0.0000] ** Chi(34)=[0.0000] ** DAVIES=[0.0000] **
---------- matrix of transition probabilities ------
Regime 1 Regime 2 Regime 3
Regime 1 0.9827 0.01725 3.898e-041
Regime 2 9.142e-015 0.9865 0.01351
Regime 3 3.872e-017 9.379e-029 1.000
\m*** Warning: The Markov chain is not ergodic!
---------- coefficients ----------------------------
neer gdp cpi wcpi
Mean (Reg.1) 116.19 19948. 104.32 96.748
Mean (Reg.2) 116.67 20122. 106.04 102.08
Mean (Reg.3) 118.56 18436. 105.20 101.31
neer_1 0.96708 1.7709 -0.025247 -0.039334
gdp_1 1.5170e-005 0.99661 3.5251e-005 7.0424e-005
cpi_1 0.038914 3.0816 0.90622 -0.078802
wcpi_1 -0.047741 -2.0045 0.047648 0.83201
SE (Reg.1) 1.3080 204.01 0.42659 1.3978
SE (Reg.2) 1.1571 389.17 0.61994 1.3593
SE (Reg.3) 1.5331 1305.3 0.70420 1.4410
---------- contemporaneous correlation -------------
Regime 1
neer gdp cpi wcpi
neer 1.0000 -0.1726 -0.2434 -0.0581
gdp -0.1726 1.0000 -0.0944 0.0090
cpi -0.2434 -0.0944 1.0000 0.2283
wcpi -0.0581 0.0090 0.2283 1.0000
Regime 2
neer gdp cpi wcpi
neer 1.0000 0.0420 0.1311 -0.0702
gdp 0.0420 1.0000 -0.1804 0.1617
cpi 0.1311 -0.1804 1.0000 0.3096
wcpi -0.0702 0.1617 0.3096 1.0000
Regime 3
neer gdp cpi wcpi
neer 1.0000 -0.0186 -0.3090 -0.1271
gdp -0.0186 1.0000 0.0351 0.1083
cpi -0.3090 0.0351 1.0000 0.2772
wcpi -0.1271 0.1083 0.2772 1.0000
*** Warning: Some transition probabilities are close to the border;
Numerical stability endangered.Runtime error: '[0 : 11][] in matrix[1][1]' invalid index range
Runtime error occurred in decomposeLambda(100), call trace:
C:\hmk\MSVAR\INCLUDE\MsvarCom.ox (100): decomposeLambda
C:\hmk\MSVAR\INCLUDE\MsvarEm.ox (419): StdErr
F:\软件安装\ox340\Ox\packages\MSVAR\96-11.OX (31): main
这是结果 那个ergodic处的警告的意思是不是不太适合我的数据啊?不具有遍历性什么含义?
您帮忙看一下结果可以吗?
如果这个模型不适合的话 我是不是要换其他的,比如说MSM–VAR,之类的呢?
还有,脉冲响应分析可以做吗?
谢谢老师,麻烦您了