Alternative unknown in dictionary

[Kara MacLeod]

Hi -

I looked at the documentation but want to confirm what these errors mean...Thanks!

Warning: Error: Alternative 36 unknown in dictionary.

Warning: Error: Alternative 54 unknown in dictionary.

Warning: Error: Alternative 0 unknown in dictionary.

[Bierlaire Michel]

This is out of context.What model did you specify? Please post the p file. 

--
You received this message because you are subscribed to the Google Groups "Biogeme" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.
To post to this group, send email to bio...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeme.
For more options, visit https://groups.google.com/d/optout.

[Huyen Le]

Hi all --

I'm having the same problem with my hybrid choice model (following the original 05latentChoiceFull.py on the website, except that I used a latent regression model for the measurement part). Below is my spec file. Prof. Bierlaire, can you diagnose this file as well? Thank you so much!

[Bierlaire Michel]

Difficult to say. It seems to be a problem with the numbering of the alternative. 

To identify the problem, I would advise to estimate first a choice model only, without latent variable. Second, to estimate the latent variable only, without the choice model. And to merge them only afterward, as described in the online tutorial. 

<LatentChoiceFull.py>

[Lizy Spanos]

Hello all! I am relatively new to using PythonBiogeme, and i wonder if this particular issue was ever resolved. I'm getting the error message:

Warning:  Error: Alternative 4 unknown in dictionary.

I'm only running a very simple MNL as well, nothing fancy! I attach the code i am using. If anyone can spot an issue in it that i didn't see, i'd appreciate the help! Been stuck on it for a few days now.

Cheers!

Lizy

[Michel Bierlaire]

It means that one of your ChoiceX variables takes value 4, while there are only 3 alternatives. 

<f6cteffects.py>

[manuel barrientos cifuentes]

Hi All
I had this trouble a lot of times, if isn't the problem that Professor Bierlaire have noticed, maybe is due the way that you saved the .dat file. I've solved it just triying different ways of generating the .dat file (finally, I'm modifiying my data in STATA, and exporting it to .txt, and after that, saving it in .dat format, probably exist a more efficient way to do it).

Best regards, 
Manuel 

[Hlulani]

Hi Everyone,

The same error surfaces in my specification, but it runs cuts the number of observations by a large number. My model only has two alternatives (1, 2), but the error ranges between "Alternative 5, 10, 15 is unknown" which are levels for one of the attributes in the model.

I thought it was an "Id" problem, so I fixed the sequence of the "Ids". I then checked the data file, but found nothing unusual (i.e. overlaps in the .dat).

At this point, I am not sure what could cause this. Attached is the _specFile.debug (how does one read this file?)

Thanks

H

[Tao Xu]

Hi everyone, 

I am working on MNL, to analyze the choice of traffic mode. This problem also happens to me. Meanwhile, there are about 441 observations in my data, but there are only 48 observations captured by the model. I have no idea what causes this issue. Attached is my mod file and issue (screen captured). Thank you very much.

Best,

Tao

[23:12:42]patFileNames.cc:48  Portland_20191229B_modechoiceMNL.par does not exist

[23:12:42]patFileNames.cc:52  Trying default.par instead

[23:12:42]patBiogeme.cc:134  Read default.par

 Opening file Portland_20191229B_modeData.dat

Warning:  Error: Alternative 2488107882 is unknown

Warning:  Error: Alternative 224 is unknown

Warning:  Error: Alternative 9 is unknown

Warning:  Error: Alternative 0 is unknown

Warning:  Error: Alternative 8 is unknown

Warning:  Error: Alternative 0 is unknown

Warning:  Error: Alternative 373 is unknown

Warning:  Error: Alternative 7 is unknown

Warning:  Error: Alternative 0 is unknown

Warning:  Error: Alternative 367 is unknown

Warning:  Error: Alternative 1536 is unknown

Warning:  Error: Alternative 515 is unknown

Warning:  Error: Alternative 0 is unknown

Warning:  Error: Alternative 41 is unknown

...............................................................

[Michel Bierlaire]

It seems that your data flle is not read properly. 

I would also advise you to use the new version of Biogeme, that is a module in Python. See the webpage biogeme.epfl.ch. 

 

-- 
You received this message because you are subscribed to the Google Groups "Biogeme" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/biogeme/f0680185-b6a5-4318-bfd4-19ed759e1c66%40googlegroups.com.
<Portland_20191229B_modechoiceMNL.mod><Issue_unknown alternative.png>