I wonder whether anybody knows how to resolve simaltaneous equations issues when running the simuation process...*!$%@ if i can get it to work...any tricks or tips?
J.
Simaltenous Equations
-
Administrator
- Super Administrator
- Posts: 4680
- Joined: Wed Mar 05, 2003 3:10 am
Simaltneous Equations
I have attached a copy of the model, it appears to the distance from the advertising saturation threshold which I have as Adverstiding Saturation threhold - Amount Adveretised, which it doesn't seem to like, the manuals suggest that it is something along the lines of an operator error, or the lack of the use fo the SImaltaneous Equation Function, both of which i have tries to some extent and nothing seems to work that effectively, may it something todo with the osicatory feed back nature of the equation.
- Attachments
-
- Market Saturation.mdl
- (3.25 KiB) Downloaded 457 times
Simaltenous Equations
Sorry about the spelling!
simultaneous equations
Hi
Your model is difficult to understand.
First, the sales should not be a level.
Second one does not see how the net profit can influence the sales. It may influence sales by influencing eventually the advertising.
Using simultaneous equations does not work, because your equations are linear. It will not work either with the find zero.
The best thing is to redraw another model with levels that represent better the dynamic of the model through the time.
Attached an example of model.
Note that there are units in the model, an absolute necessity in any serious model.
In the model you should replace the effects equations by table functions, because all these effects are not linear.
The linearity of the model explains the very odd results.
The parameters should too be considered with care.
I cannot build the table functions nor give values to the parameters not knowing your problem.
joined the model transformed.
Regards.
JJ
Your model is difficult to understand.
First, the sales should not be a level.
Second one does not see how the net profit can influence the sales. It may influence sales by influencing eventually the advertising.
Using simultaneous equations does not work, because your equations are linear. It will not work either with the find zero.
The best thing is to redraw another model with levels that represent better the dynamic of the model through the time.
Attached an example of model.
Note that there are units in the model, an absolute necessity in any serious model.
In the model you should replace the effects equations by table functions, because all these effects are not linear.
The linearity of the model explains the very odd results.
The parameters should too be considered with care.
I cannot build the table functions nor give values to the parameters not knowing your problem.
joined the model transformed.
Regards.
JJ
- Attachments
-
- Market Saturation_2.mdl
- (7.85 KiB) Downloaded 445 times
simultaneous equations
Hi
Another modified model joined, where the influence from advertising has been erased as it was in the initial model and where the behaviour is more possible and stabilizes after a while.
I have too modified the values of the parameters.
The effects of both saturations should be better represented by a 2 dimension lookup explained in the definition of the vector lookup of Vensim.
Regards.
JJ
Another modified model joined, where the influence from advertising has been erased as it was in the initial model and where the behaviour is more possible and stabilizes after a while.
I have too modified the values of the parameters.
The effects of both saturations should be better represented by a 2 dimension lookup explained in the definition of the vector lookup of Vensim.
Regards.
JJ
- Attachments
-
- Market Saturation_3.mdl
- (7.49 KiB) Downloaded 463 times