conclusions from the results of the model, we compare our explanation of the events
with the reality of such events, and finally we advance our understanding, refine our
questions, and improve our models. If we are really good, we make and test
predictions with the model. According to John L. Casti, good models are the
simplest that explain most of the data from an operating system, and yet do not
explain it all, leaving some room for the model, or theory, to grow [3]. Good models
must have elements that directly correspond to objects in the real world. All models
are necessarily simple constructs of reality; some are just too simple, some just
too complex.
The biological modeling process is certainly a knowledge capturing one and yet
it is exceedingly difficult to do well. Nature seems ineffable. The ultimate difficulty
stems from the complexity of the living system, both its structure and its dynamics,
and the impossibility of removing ourselves fully to observational status. Even an
expert in biochemistry, in medicine or in ecosystem analysis cannot make much
headway against such complexity alone. A team of such people, expert in the
various aspects of a problem may be the answer. As a team, researchers can more
easily see anthropomorphism in each other. We suspect that such team modeling
will become much more common in the near future. The elements that make such a
team successful are of course the possession of real and pertinent expertise,
compatible personalities, a common concept of the questions to be answered, and
a common modeling language. Since the level of modeling expertise is likely to
vary among even the best set of experts, a simple modeling language is needed—
one that can be understood by each of the team in a very short time. The programming language STELLA meets that requirement. Our book contains the guidelines
for the most likely successful process of team or individual modeling and it is based
on that language. Many generations of students both at universities, in government,
and in corporations have contributed to the process described herein.
1.2 Static, Comparative Static, and Dynamic Models
Three general types of models can be defined. The first type of models is those that
represent a particular phenomenon at a point of time, a static model. For example, a
map of the world might show the location and size of a city or the location of
infection with a particular disease, each in a given year. The second type comprises
comparative static models that compare some phenomena at different points in
time. This is like using a series of snapshots to make inferences about the system’s
path from one point in time to another without modeling that process itself.
Other models describe and analyze the very processes by which a particular
phenomenon is created. We may develop a mathematical model describing the
change in the rate of migration to or from a city, or the change in the rate of
the spreading of a disease. Similarly, we may develop a model that represents the
change of these rates over time. This latter type of models is dynamic in the sense
that they attempt to capture the change in real or simulated time.
6
1 Modeling Dynamic Biological Systems
with the reality of such events, and finally we advance our understanding, refine our
questions, and improve our models. If we are really good, we make and test
predictions with the model. According to John L. Casti, good models are the
simplest that explain most of the data from an operating system, and yet do not
explain it all, leaving some room for the model, or theory, to grow [3]. Good models
must have elements that directly correspond to objects in the real world. All models
are necessarily simple constructs of reality; some are just too simple, some just
too complex.
The biological modeling process is certainly a knowledge capturing one and yet
it is exceedingly difficult to do well. Nature seems ineffable. The ultimate difficulty
stems from the complexity of the living system, both its structure and its dynamics,
and the impossibility of removing ourselves fully to observational status. Even an
expert in biochemistry, in medicine or in ecosystem analysis cannot make much
headway against such complexity alone. A team of such people, expert in the
various aspects of a problem may be the answer. As a team, researchers can more
easily see anthropomorphism in each other. We suspect that such team modeling
will become much more common in the near future. The elements that make such a
team successful are of course the possession of real and pertinent expertise,
compatible personalities, a common concept of the questions to be answered, and
a common modeling language. Since the level of modeling expertise is likely to
vary among even the best set of experts, a simple modeling language is needed—
one that can be understood by each of the team in a very short time. The programming language STELLA meets that requirement. Our book contains the guidelines
for the most likely successful process of team or individual modeling and it is based
on that language. Many generations of students both at universities, in government,
and in corporations have contributed to the process described herein.
1.2 Static, Comparative Static, and Dynamic Models
Three general types of models can be defined. The first type of models is those that
represent a particular phenomenon at a point of time, a static model. For example, a
map of the world might show the location and size of a city or the location of
infection with a particular disease, each in a given year. The second type comprises
comparative static models that compare some phenomena at different points in
time. This is like using a series of snapshots to make inferences about the system’s
path from one point in time to another without modeling that process itself.
Other models describe and analyze the very processes by which a particular
phenomenon is created. We may develop a mathematical model describing the
change in the rate of migration to or from a city, or the change in the rate of
the spreading of a disease. Similarly, we may develop a model that represents the
change of these rates over time. This latter type of models is dynamic in the sense
that they attempt to capture the change in real or simulated time.
6
1 Modeling Dynamic Biological Systems