http://qs321.pair.com?node_id=574603

lin0 has asked for the wisdom of the Perl Monks concerning the following question:

Greetings,

I am searching for Granular Computing implementations using Perl but I have not found anything, yet. Does anyone here have any pointer to such implementations?

On my side, I started porting code I had written in other languages to Perl. The first piece I did was a Fuzzy C-Means implementation. I posted it on lin0's scratchpad. Any comment on how to improve it would be welcome. Now, the thing is that I do not want to spend time writing code that is already available. So, if anyone has already written some granular computing code and were willing to share it, I would love to have a look at it. Maybe, we could work together to implement additional functions (or why not, to write a CPAN module). If there is nothing out there, I just keep porting my own code.

Thank you,

lin0

Replies are listed 'Best First'.
Re: Searching for information about Granular Computing with Perl
by jkeenan1 (Deacon) on Sep 25, 2006 at 01:01 UTC
    1. Googling for 'granular computing Perl' didn't turn up anything obviously pertinent, so I guess you're as welcome as anyone to introduce granular computing to the Perl community (or at least to the monastery).

    2. The posting on your scratchpad would be a bit more useful if, instead of expecting to get the data via a handle to a file, you were to supply us with a dozen or so lines of sample data under a __DATA__ tag at the end of the script.

    Jim Keenan

      Hi Jim,

      Thank you for your comments.

      About your first comment, do you mean that I should write a small tutorial about Granular Computing to introduce it to our fellow Monks? Or do you simply mean that I should keep writing the code and ask for feedback from our community? For the former case, I could certainly start by describing what the Fuzzy C-means does and how it can be used in Granular Computing. It will take me a couple of weeks, though, because I am quite busy right now with other stuff. However, to give you a preview, I can tell you that what the Fuzzy C-Means does is to look for groups in the data (that is, it groups patterns according to their similarity. This is why, some people say that the Fuzzy C-means searches for structure in the data). For the latter case, that is what I am doing: porting my code to Perl. When I get more code written, I will certainly post it on the Monastery to ask for feedback.

      About your second comment, I just posted some sample data as you suggested. It is a very simple dataset. If you want to play with it, what I suggest you do is the following:

      1. Plot the data using your favourite program. I suggest you use an XY chart. On this chart, you might notice that there are two groups of patterns: one close to (0, 0) and the other close to (12, 10)
      2. Run the code available in the lin0's scratchpad without any argument (without arguments it will search for two groups). One of the groups will have a prototype (sort of representative element) about (1.0, 0.9) while the second one will have a prototype about (11.5, 11.2)

      I really was not expecting to discuss much about the code but if people are interested, I could certainly write a node describing it in more detail.

      Thanks again,

      lin0

        "About your first comment, do you mean that I should write a small tutorial about Granular Computing to introduce it to our fellow Monks?"

        I can't speak for Jim, but I personally would welcome a short introductory tutorial, with short introductory examples in perl.

        Is the Wikipedia entry on Granular Computing accurate?