I've been liking Gretl, It is simple, light, fast, free. Ir has been making by an Economic professor.

http://gretl.sourceforge.net

Thank you Hermann.  I will look at the Gretl tool.  Does it run on Ubuntu platforms easily?  You may be interested to follow the debate at this question "How do the economists define self-interest and rationality? "  You will then see why I am looking for an 'advanced' tool.....  and probably a human driver also!!

Dear Mohammad, try correlation analysis from econometry.

Dear Rogovska

Please, address your reply to Sir Robin.

Heariest Thanks

Israr

Dear Vanda,

My degree, admittedly only at BA level, is one of Econometrics with specialisation in multivariate correlation and I-O Analysis using matrix algebra and linear programming.  I find that although these techniques are suited to identifying and exploiting correlations between time series of data points so as to provide a descriptive analysis of 'state' snapshots along a time line.

I did not make myself very clear in my question for the sake of brevity.  The need is, in fact complex.  I do not wish to only discover significant correlations by building a matrix of independent equations.  This is part of the preparatory work which would be necessary, it is true.  My goal is to build a modelling system which is essentially a simulation exercise. One which can follow the interactions of a set of independent agents which are themselves described by sets of interrelated behavioural functions the output of which at one point in time feeds into the next.

The difficulty is to stop short of designing it as a 'hard problem' which cannot be practicably computed, but sufficiently complex that it can be realistic enough to have useful output.  And to be a system of computing that is sufficiently general that it can be applied across sets of interrelated groups of sub-problems so that the 'hard problem' difficulty can be avoided as the problem scope is scaled to the real world need.  I might also add that I am a real world computer system engineer and analyst.  I can see that this task is as much one of mathematical tools as it of computational process design.  But it needs the skills and knowledge of mathematical experts working together with subject theorists to design a pragmatic approach.

The tool set you mention does not seem to me to be capable of providing a suitable mathematical framework for modelling multiple interacting 'processes' which are taking place between multiple agents with potentially independent goals along a time line.  Such that the inter-agent interactions that are effected in between one 'point of state' and the next are not a finite set but can themselves be subject to the effects generated by point values contained in multiple 'time series of externalities', each point of which  may have a different resultant level of 'effect' upon the 'state' value established for each next 'point of state' of any of the behaviour functions of each individual agent.  The coefficients in these agent behaviour functions would indeed most likely be determined by multivariate correlation analyses.  However, the particular coefficient value applied to a particular B-function of a particular agent would be allocated according to a desired distribution across the agents from specific coefficient values drawn from sampling the set of values representative of the probability curve characteristic of the 'best value' estimate for the coefficient which the behavioural correlation analysis stage has produced.

I don't know if that is clear, or helpful to you.  Israr knows too well how convoluted my thinking can get!

Best regards,

Robin

For the possible benefit of others who may be reading this.

I have found that Evolutionary Economics focusses on the development of computational designs which are useful to the resolution of difficult mathematical requirements.  Notably this includes the passing of derived values from calculations made at one point in time to recalculations made at the next.  This can allow for the inclusion and reading out of variables and parameters in between each stage.  They are , thus, simulation designs.

Refer to the work of Yoshinori  Shiozawa at Osaka.

Try cellular automata

https://www.researchgate.net/publication/265253105_Tourism_and_Simulacrum_The_Computational_Economy_of_Algorithmic_Destinations

Article Tourism and Simulacrum: The Computational Economy of Algorit...

Dear Mr. Lew,  Thank you for your suggestion.  I will study your paper to see how the approaches might be adaptable to the problem environment I would like to examine.

Best regards,

Robin Jarvis

suitable mathematical tools  for modelling intra-community economic production flows over several periods are the covariance model (econometric model with dummy variables which include the interaction effects between the quantitative and qualitative variables)

Dear Bernard,  I have read your paper with interest.  I must commend you on a very precise and 'tight' paper.  Every word was loaded with 'significant meaning'.  Especially in their philosophical implications.  These are very close to my own thinking. 

There is a small difference between my intended use of computational tools and their usual focus.  And it arises, as I can now see, from what you have identified so clearly as the AC and IC transformation.  I have recently written in discussions in RG of the way in which the philosophical framework of Mercantile Europe gave rise to economic models in which were embedded the simplified, virtualisation of that moral and societal paradigm.  These 'game models' have guided and justified the creation of the complexity of institutional processes and structures which now 'create' our present reality.  They are the 'rules of the game' per the neoclassical simplification.  The AC of Adam Smith and Ricardo have become the basis of our IC state.  Our economies have already been gamified.

This has many consequences, two are very major.  The first is, as you have cited, 'complexity is resilient' and so necessarily the opposite is true.  Simplification is Fragile.  Globalisation, monopoly and nation states are simplifying structures and tend towards a system with increasingly low resilience at a time when we have become acutely aware of the need for greater resilience.

Second, by their nature economic structures, processes, agreements and contracts are the 'rules' and 'bounds' or 'constraints' contributing to the achievement of a maximal production goal. Production has the attribute of ownership and that denotes wealth. Wealth is the basis of the 'scoring' mechanism. In reality it is also the attribute that imbues some game players with more power, and therefore more influence, than others in the evolution and 'learning/developmental' aspects of the game.  They can and do write in new rules and protect any that benefit them and their 'neighbours'.  This includes the creation of 'institutions' which do it for them and which are given their own goals of resisting any changes from any sources other than their own.

However, any change for the good at the micro- and macro- economic level of economic relations across the world, as you have so clearly identified, does require necessarily and quite urgently a change in the rules of the game.

Therefore my use of computational modelling is not to collect data and insert it into a simulation model to examine and improve/tweak life as it is.  I hope to grow and develop new models of new realities with an objective function of universal well-being rather than one of production maximisation which what the current metaphysics would have us believe in.

I had paused in that endeavour due to the doubts that my background in covariant and input-output modelling, from many years ago, raised in my mind as to the suitability of those tools and their roots in describing what is rather than in questioning what might be.

Thank you once again for your paper!  As I read it I was reminded of other seminal works that have had very great impact upon our present reality.  I repeat here that, in my very humble opinion, it has implications wider than you imagine and the ideas you have so successfully brought together into a single narrative should be required reading in every academic discipline.  It brings ideas of social psychology and social anthropology across 'the line' into the significant business of solving the problem of how we actually get to survive and live our daily lives.  Ideas such as the 'social construction of reality' and the psychology of cognition wherein interpretation of external stimuli into a consistent 'reality' requires the inclusion of a 'metaphorical framework' by which the stimuli can be understood and caused to make sense to us.

I am sure that we shall talk again....

Best regards,  Robin

Prof Soliman,  I owe you a debt of gratitude for refreshing my ancient (1975)  memories!  Also an apology for not thanking you earlier.  I have been away for a few months and I am only now settling back into 'thinking' mode. 

Thank you for your mention of how dummy variables in covariance models are used.  I had completely forgotten.   It is a long time since economic modelling was my focus.  However I cannot say the same is true for the use of binary state indicators in computer programme design. I should have made the connection.  Thank you