15 Reputation

2 Badges

10 years, 86 days

MaplePrimes Activity

These are answers submitted by Michal

So I'm coming from  mathematical economics background. And here are some important things that would have to be answered for your model design:

  1. Are you wanting a hypothetical model or a real world model? It seems to me that you are wanting a model to enforce some real world ideal, but that ideal may not be realistic. The law of comparative advantage is likely the most popular way to counter your claim, but not the only one. For example, you mention that the majority of spending to local businesses stays within the community, but the line between communities is entirely arbitrary. If I view things as a global community, then 100% of all money spent anywhere stays within the community. Furthermore, you don't question whether or not big businesses bring money into a community from outside sources. Nor do you mention whether employment increases, how much additional money is available to members iff goods at a large business are cheaper than at local business (this is an aspect of economies of scale). Now, if you are wanting a hypothetical model not based on real world scenarios, I see no problem here. In fact, you don't need any facts -- you simply need the model design and then you find what characteristics the model has. Doing the first, however, is unscientific as it is ignoring evidence to the contrary.. 
  2. Some of your "facts" seem faulty... e.g. "local businesses rely on local services" ... Having both worked in local businesses, national chains, and then studying both while earning my degree, I can assure you that both groups use both local and national supplies. For one, they all use local labor (one of the two inputs for output in local equations). Second, they often use manufacturing facilities that are local for the goods that are convenient to do so (think produce). And again a lot of these local/non-local distinctions are arbitrary. As are the big/small business distinctions.

So if you are looking for a hypothetical model, that could be interesting. My first thought falls towards some form of a system of differential equations such as Loetka Volterra predator-prey. A second way might be to show that as such large entities collapse, it is hard to replace them (in practice the only thing that actually causes them to collapse is replacement -- Think RIM vs iPhone/Android). If I were designing such a model this is where I would start. 

So grid is meant for a cluster of computers not a multiple processors... 

http://www.maplesoft.com/products/toolboxes/gridcomputing/index.aspx see here for grid


What you're wanting is threads http://www.maplesoft.com/support/help/Maple/view.aspx?path=Threads see there


Note that the second link even offers examples.


I do have a question, however... what kind of program is this? 50K for loop is gonna be intensive in maple. I would be far more worried the loop would max the ram if you are running multiple computations in the loop that require a large number of variables... if the computations are light... maybe not so bad... basically I am thinking about this... 

Honestly, I'm rather new to maple so I'm not quite sure on how much it will optimize... but it is an interpreted language. Even C++ as highly optimized as it is would get angry if you were doing something like adding an element to an array 50K times. I might first attempt to find an alternative method of computation before I worry about the threading.

Page 1 of 1