ETS, not smooth to the end time point

For notifying us of what you believe are bugs or errors in EViews.
Please ensure your copy of EViews is up-to-date before posting.

Moderators: EViews Gareth, EViews Moderator

maxchen
Posts: 191
Joined: Fri Oct 10, 2008 4:03 pm

ETS, not smooth to the end time point

Postby maxchen » Mon Jun 29, 2015 1:38 am

EViews8Patch_042715

ets.7z
Test-ets.prg and workfile (extract from a large project)
(409.25 KiB) Downloaded 256 times


run the prg file Test-ets.prg (modify %path in line 9 if needed), when the prg stops
1 osv(3) = "J2012", and the smooth series COCS, smooth to 1/20/2012 only, not to the end date 2/22/2012 (in osv(2) )
2 if uncomment line 35 (!oi=140), when prg stops, osv(3) = "Q2012". which means ETS smooth to the end date on contract J2012 (fail when !oi=139), but not to end date on contract Q2012

I feel that there is some random behaviour in ETS, is it a bug?

Let's have another try: uncomment line 81 (pagesave) in Test-ets.prg, the active workfile page is extracted. Then run the following code, they forecast to the end date. However, if you uncomment line 83-85 of Test-ets.prg to do the same job, they fail.

Code: Select all

!T = 60   '# of obs in estimation
%otb1 = "1/20/2012"
%otc = "2/22/2012"

' wfopen J2012-1m2.wf1
wfopen Q2012-1m2.wf1

delete cocs
smpl %otb1-!T+1 %otb1
freeze(mode=overwrite, ootb) coc.ets(e=e, t=e, s=e, modsel=hq, forc=%otc) cocs
scalar ets1 = @elem(cocs,%otc)



one more question: if using GUI, the ETS has an option "Reject non-optimized models", how to active this option when running by cmd line?

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13317
Joined: Tue Sep 16, 2008 5:38 pm

Re: ETS, not smooth to the end time point

Postby EViews Gareth » Mon Jun 29, 2015 3:07 pm

I've spent some time trying to decipher your post, but I'm afraid I haven't got very far. I can't see what you're describing.

I will note that there isn't, currently, a command line version of the non-converge model option. It will be added (noconv) in the next ev9 patch.
Follow us on Twitter @IHSEViews

maxchen
Posts: 191
Joined: Fri Oct 10, 2008 4:03 pm

Re: ETS, not smooth to the end time point

Postby maxchen » Mon Jun 29, 2015 6:57 pm

Thanks for the effort.

to reproduce the scene, I capture the screen as a video on http://youtu.be/E9iPzv4BgyA

maxchen
Posts: 191
Joined: Fri Oct 10, 2008 4:03 pm

Re: ETS, not smooth to the end time point

Postby maxchen » Tue Jun 30, 2015 5:27 am

I test in EV 9, the same (wrong) results
EV9-T.png
runing on EViews 9
EV9-T.png (79.12 KiB) Viewed 4023 times

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13317
Joined: Tue Sep 16, 2008 5:38 pm

Re: ETS, not smooth to the end time point

Postby EViews Gareth » Tue Jun 30, 2015 8:39 am

Got it, thanks.

Bug fix will be posted in next EV9 patch. The bug was due to a machine error calculation causing a zero to actually be -1e-30 (i.e. a very tiny negative number), which then caused NAs to be generated.
Follow us on Twitter @IHSEViews


Return to “Bug Reports”

Who is online

Users browsing this forum: No registered users and 22 guests