ucsd-garch工具包及其正确安装方法 [推广有奖]

31楼

gssdzc(真实交易用户)

发表于 2011-8-16 21:47:29

thanks a lot

32楼

caoguangxi(真实交易用户) 发表于 2012-1-14 05:52:35

我的重新安装后怎么出现以下问题了阿, 请高手帮助！
We can now test to see if the residuals are conditionally normal using kolmogorov as detailed earlier.
We call this test by [statistic, siglevel, H]=kolmorgorov(errors./sqrt(ht),.05,'norm_cdf').
The calculated statistic 0.0557 was with a sig level of 0.0000.
Thus we reject the null that the model is adequate.
We can now see if a more complicated variance estimation will be sufficient, the Asymetric Power Garch.
We can call this type of GARCH using the multigarch function.
We call this by: [parametersAP, likelihoodAP, stderrorsAP, robustSEAP, htAP, scoresAP]=multigarch(errors,1,1,'APGARCH','NORMAL')

Press any key to estimate.??? Error using ==> fmincon
FMINCON cannot continue because user supplied objective function failed with the following error:
Undefined command/function 'tarchcore'.

Error in ==> tarch at 149
[parameters, LLF, EXITFLAG, OUTPUT, LAMBDA, GRAD] = fmincon('tarchlikelihood', [omega ; alpha; tarch ; beta] ,sumA , sumB ,[] , [] , LB , UB,[],options,data, p ,o, q, m, stdEstimate, type);

Error in ==> multigarch at 157
startingvalues(1:1+p+q+o)=tarch(data,p,o,q,'TARCH',[],newoptions);

Error in ==> ucsd_garch_demo at 196
[parametersAP, likelihoodAP, stderrorsAP, robustSEAP, htAP, scoresAP]=multigarch(errors,1,1,1,'APGARCH','NORMAL');