www.modeler-forum.com

Hello, I am wondering how to revert to qualitative modeling once I have clicked on "Input" or "Formula". Also, I am wondering what the button "normal factor" stands for, as it is always inactive.. Thanks! Grégoire.

If you want to go back to quantitative modeling, please just click the properties button at the bottom.

At the moment "Normal factor" is the only option. Because of that the button is disabled. In future when the Process iMODELER is completed there will be also Process iMODELER factor types selectable.

Magnus from Consideo

Thanks Magnus! Actually it seems that once you click on "formula" and write something in it, iModeler gets a bit lost: if I delete everything in the formula, I have an error message about validation... perhaps an extra button (qualitative/quantitative) to trigger one or the other mode would be useful...

Is this a bug?:

in quantitative mode, the insight-matrix depends on the "qualitative" weights of the influences, not on the influences that result from formulas...

example: rainfall --0.17--> pasture--0.17-->cattle

gives the same insight-matrix than

rainfall: 100;
pasture: if(time()==0;1;valuebefore(0) +delay0(1; [Rainfall]))
cattle: if(time()==0;0;valuebefore(0) +delay0(1; [Pasture]))
provided the weights in the arrows in the qualitative model are the same (i.e. 0.17; 0.17)

If I change the weights of the arrows, I change the insight matrix even if I don't touch the formulas. Apparently the insight matrix is independent of the formulas, which is misleading... I expected that it resulted from simulations of the quantitative model!

Thanks in advance for your help!
Grégoire

That's a very important post! Qualitative models answer the question, what is more or less effective, whether its a target, a measure, a risk etc.. They don't show to what extent something develops or influences. A matrix just says that on factor has more or less impact compared to other factors and that these influences may change over time.
Quantitative models show how and to what extent something develops over time. The result is a time series.
Here is a quick guide for this:
http://www.consideo.de/MODELSTDu/Englis ... re_iM.html

PS:
FCM btw also show just the resonance of an impulse - not the development of a factor.

_________________
Kai Neumann

Consideo