Jump to content


Photo

Error message in NLME model-Index out of range


  • Please log in to reply
13 replies to this topic

#1 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 07:28 AM

Hi

I keep getting the follwoing error message while running my model:

"Index was out of range. Must be non-negative and less than the size of the collection.

Parameter name: index".

Has anyone had this before and know how to fix it ?

many thanks,

Joanna


  • Seo-Ul-tix likes this

#2 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 07:32 AM

Dear Joanna

Can you send me the project or at least the model text and mapping you used. Isaw this error a long time ago but I do nto remember what was the cause other than a model error (or mapping error).

best

Serge



#3 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 07:45 AM

Hi Serge

thanks for a quick reply - I have attached a model code, and also a screenshot of the mapping, "input option" tab and also the "Initial estimates" tab. From the "initial estimates" tab it looks like the profile is generated but I could not run it without getting the error message.

many thanks,

Joanna



#4 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 07:48 AM

I do not see any attachement.

best

Serge



#5 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 07:49 AM

can you send the attachement at poppharm@gmail.com? I will check in 2 hours the status but this error is familiar.

best

Serge



#6 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 07:50 AM

I think there was an error while I sent it last time - I try again now (it should be two files)



#7 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 07:55 AM

Hi

I have just sent the attachements to the email address. It looks like there was some problem with sending them through the forum,

many thanks,

Joanna



#8 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 09:12 AM

Dear Joanna

I found the problem. It is the freeze statement for omega that had a bug.

The bug has been fixed but you do not have the most recent version of the program. In the next release freeze with omega and edit as textual will work.

Best

Serge

 

PS: Since you are using a 0 variance in one of the freeze statements, just change your code and remove that random effect and delete that freeze statement. for the other omega that ahs freeze variance, remove the freeze option. Then you should not get that error anymore.

best Regards;

Serge



#9 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 09:43 AM

Hi Serge

great - thanks for looking into the code. I have now removed the freeze variance statement and I'm re-running the model. I hope it works OK now but it may take a while to find out - last time it run for 6.5 hours before I got the Index error message, hopefully now it will run without problems,

thanks again,

Joanna



#10 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 12:07 PM

When it crashed, apparenlty it was the end of the run where the program needs to deliver all the outputs. As far as I know the freeze problem was only linked to variances and not fixed effects (you have a freeze fixed effect too). Therefore I think this crash should not show up.

If it is agin then we will deal with it.What engine are you using for optimization (FOCE ELS, QRPEM,?)

best

Serge

 

Moderators Note - I've added the files, I had to change the extension to TXT to be acceptable to the forum. Sorry. Simon. [file name=NMLE_code_fixed.txt size=1470]http://pharsight.com/extranet/media/kunena/attachments/legacy/files/NMLE_code_fixed.txt[/file] [file name=NMLE_code.txt size=1516]http://pharsight.com/extranet/media/kunena/attachments/legacy/files/NMLE_code.txt[/file]



#11 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 18 October 2012 - 12:36 PM

It worked this time - looks like freezing eta was the problem here so many thanks for pointing this out. Also, does it mean that sigma also cannot be fixed (in this version of Phoenix) ?

At the moment I only run the model using naive-pooled method just to get it to work but I will try with FOCE.

many thanks,

Joanna



#12 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 18 October 2012 - 01:19 PM

Dear Joanna

You can freeze epsilon. NO issue there.

best Regards;

Serge



#13 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,318 posts

Posted 19 October 2012 - 03:42 PM

Joanna, I just wanted to clarify that this issue was noted in the Phoenix Notification No. 14 that can be found on the Support site;

PHOENIX NLME DEFECT:
Incorrect operation when text model is executed in edit mode and Omegas are frozen (QC 12067): When a text model is run with a frozen omega matrix where the freeze was entered in text mode, it completes with the correct answer (happens with all engines), but then fails upon returning to the UI with an 'index out of range' error.

On the other hand, this does not occur if the model is a library (built in) model and the omega freeze is implemented by check box.

The only known workaround if the user wants to execute the text model with that option is to execute the model in the Command-Line mode.


Also you will see I have posted the corrected PML files above to Serge's mail. Out of interest are you utilising the HPC Grid you have at AZ for these runs? How is that working for you?

Simon.

#14 Joanna Parkinson

Joanna Parkinson

    Newbie

  • Members
  • Pip
  • 7 posts

Posted 23 October 2012 - 01:44 PM

Hi Simon

thanks for letting me know. The model seems to be working OK now, once I removed the fixed omega.

Also, I'm not actually using the HPC Grid myself, so I can't comment on that..

Joanna






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users