simultaneous equations

[Panichel]

Hi,
I have a problem in my model. When trying to simulate the model I get an error of simultaneous equations and I cant solve it !!
I have check vensium documentation in this error but I cant solve it !
I post the model. Can somebody take a look please?
Thanks !!!

[LAUJJL]

Hi

joined a corrected model that works with the simulataneous function. But there is still a problem with the calculation of a variable. I have added a zidz funtion to avoid dividing by zero. It does not prove that it is a valid correction. It is maybe necessary to add a level somewhere it it is not manadatory to loop within a time step. I have initialized the simultaneous equation to 0.
With the initial value of 0, the price values are 0 which does not make sense. I have tried positive and negative values. they do not converge. If you want absolutely to use a loop and know why, you can try the find zero function that provides more control on the loop.
The demand depends too linearly from the price and the elasticity which is never the case, only on a small intervals.
You must correct too the unit error and maybe too the warning that shows a function table not normalized (better practice).
Regards.
JJ

[Edited on 6-2-2010 by LAUJJL]

[Edited on 6-2-2010 by LAUJJL]

[Edited on 6-2-2010 by LAUJJL]

[bob@vensim.com]

I would recommend against using a simultaneous construct for this problem. Price is almost always most realistically treated as a level. The level can be adjusted up or down based on market conditions. See, for example, the smooth pricing molecule.

[Panichel]

Hi,
Thanks for the correction. Actually, I have tried to reproduce Meadows commodity cycle, where prices are auxiliary variables and there is a feedback loop on consumption-inventory coverage and prices. But I dont know why I get this simultaneous equiation error and Meadows dont. Can I solve this problem by seting price as a level?
Thanks again for your help
Luis

[LAUJJL]

Hi

Just changing the price into a level and hoping that it will solve the problem is not necessary the solution.
You must understand how the model works in relation to the problem you try to solve. You can always try and it may work by chance.
Otherwise you may formulate a written description of the problem with simple english words to see the relation between the model you build and the written formulation. It is difficult to help anybody on a problem where the only description of the problem is through a finished model eventually with conceptual or technical bugs.
JJ

[Edited on 7-2-2010 by LAUJJL]

[bob@vensim.com]

In this case price is indeed directly driven by inventory - but there is s smooth of that (to expected hog price) that is used in determining supply decisions and another level for determining demand. I am attaching a version of this model in Vensim that will probably be helpful.