Page 1 of 1

EViews seems to crash randomly (when doing a quantile regression)

Posted: Wed Mar 28, 2018 7:57 am
by tho_mi
Hey,

I'm not sure whether this is the correct "area" to post this, but it seems most suitable. Colleagues of mine have problems with a script and asked me to fix it because EViews crashed when they ran it. I tried to figure out where the problem is. The first few times I ran the script EViews crashed when doing a quantile regression (with parameters quant=0.1, cov=boot, btmethod=mcmba). My first suspicion was the Bootstrap method. I tried different ones and it seemed to work. "Unfortunately" the original settings now work too. Could there be any reason why EViews crashed first and seems not to anymore?
Additionally, if EViews crashes while doing a certain kind of regression, how can I find out what the exact problem was?

Thanks in advance!

Best,
Thomas

edit:
I'm using EViews 9.5 and I'm not sure which version they are using.

Re: EViews seems to crash randomly (when doing a quantile regression)

Posted: Wed Mar 28, 2018 8:04 am
by EViews Gareth
Best thing is to email support@eviews.com with the program and data.

Re: EViews seems to crash randomly (when doing a quantile regression)

Posted: Wed Mar 28, 2018 8:08 am
by tho_mi
EViews Gareth wrote:Best thing is to email support@eviews.com with the program and data.


It's urgent and I'm not sure whether I'm allowed to give the data to someone. Do you have any idea what I could try?

edit:
I ran the script on a different computer. Once I change the bootstrap method to "mcmb" instead of "mcmba" and run the line of code that leads to the problem once with that setting, the original code works just fine.

edit2:
I restarted the computer and EViews crashes again. "Solution" still works. I assume it has to do with a temporary file or so that gets deleted after a restart (restarting EViews has no effect).

edit3:
At least in my case it makes more sense to change the bootstrap method from "mcmba" to "xypair" ("mcmb" leads to strange results, probably due to autocorrelation).

Re: EViews seems to crash randomly (when doing a quantile regression)

Posted: Wed Mar 28, 2018 8:12 am
by startz
One small thing to do is to be sure that your copy of EViews is up-to-date.