Re: Sampling from a simplex: part 2
- To: mathgroup at smc.vnet.net
- Subject: [mg95904] Re: Sampling from a simplex: part 2
- From: Andreas <aagas at ix.netcom.com>
- Date: Thu, 29 Jan 2009 05:58:00 -0500 (EST)
The following from Wolfram Tech Support yields the same results as mine and simplifies the code:
sampleSimplex[sampleSize_, dimensions_, min_, max_] :=
Module[{s, tableNormal},
s = RandomReal[ExponentialDistribution[2], {sampleSize, dimensions}];
Map[min + Normalize[#, Total]*(max - min) &, s]]
Can anyone confirm that this will indeed give me the kind of sample I need i.e. where min = -1 and max = 2?
Also, this still leaves me with the larger question of approaches or rules of thumb to refactor Mathematica code (along the lines of Martin Fowler's book "Refactoring - Improving The Design of Existing Code" targeted predominately at object oriented languages) or simply to write the most concise, clear and robust code.
Again, any suggestions, guidelines, or resource suggestions along these lines much appreciated.