in reply to Re^3: Fractal structure: PDL, memory, time
in thread Fractal structure: PDL, memory, time

Hi BrowserUK,

This code is fanstastic! I am impressed.

I hope to be able to fit it to my needs (indices), which I've been dealing with so far by nesting structures, and not completely satisfied with the runtime, maybe this code will be faster, or something inspired by it...

Frank :-)

  • Comment on Re^4: Fractal structure: PDL, memory, time

Replies are listed 'Best First'.
Re^5: Fractal structure: PDL, memory, time
by BrowserUk (Patriarch) on Jul 25, 2007 at 18:47 UTC

    I went back to your original post to remind myself of what you are doing with the indices...and there it was NOT :)

    It was something to do with ordering the indices for a row, stabley, descending by level value in the cell?


    Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
    "Science is about questioning the status quo. Questioning authority".
    In the absence of evidence, opinion is indistinguishable from prejudice.
      Hi BrowserUK,

      The question was: out of that fractal structure given depth & line index:

      Output indices for (let's call this level) $level = $max, $level = $max-1, $level = $max -2 untill some condition has been met. In my environment, this condition will be met at level 0, or 1 or 2 in most cases, so checking the indice for low levels without building too much may be faster.

      I've done this by two different methods. The first one involves building the whole structure for each line, and then goes into it and checks indices.

      The second one involves having lines as base 4 indexes into a nested structure, then look at the closest ones. For that second method (possibly faster) I've used both Algorithm::Loops & Math::Combinatorics.

      Yet I am not happy with runtime... I'm curious whether this can be optimized using the fractal structure and your code.

      :-)

      Frank