models are not open source, it is usually an 'all or nothing' deal: either you use the model as is, or drop it entirely if it does not have some of the features needed for your study.
Some models are intentionally designed as games, with special emphasis put on the graphic interface and ease of use. One good example is the SimCity computer game, which has a sophisticated socioeconomic and ecological model at its core, but no one ever saw this model and rarely questioned whether it can be calibrated or validated. The purpose here is to enhance the interactive utility of the program, to maximize its user friendliness, and simplify the learning process.
Unlike prefabricated models, which are after all developed for specific systems, there are also generic software tools that can help build models of any systems. These are probably most interesting to consider when a particular modeling task is in order. The problem with choosing an appropriate tool is a difficult one, because learning each one of them requires some time and effort, which can be quite considerable. Therefore, more often we see that once modelers master a particular modeling language or system, they are inclined to use the same acquired skills next time they need to analyze a different system, even when this other system is quite unlike to the first one, and even when the modeling goals are different. As in the well-known saying, apparently by Bernard Baruch (or, according to other sources, Abraham Maslow), ''If all you have is a hammer, everything looks like a nail.'' Anyway, it is quite natural to try to do what you already know how to do. As a result, modelers who would be equally proficient in a variety of modeling techniques are quite rare, and good comparisons of modeling tools are also hard to find.
Here we will give a brief overview of some software tools that are available for modeling with some recommendations about their applicability. Note that neither of these tools implies any kind of core model - they can be used to put together any model. However, each one assumes a particular modeling paradigm and therefore has certain limitations.
Was this article helpful?