in reply to Re^4: Pattern enumeration.
in thread Pattern enumeration.

That was for the first version, the later one runs much faster. Actually it took it 2600 minutes to solve the 8x8,6 case:
$ time /tmp/pc 8 6 the number of patterns is: 00000003e2e2292feeb6f4b8019b3e59cd564cb95ff +f38e0 real 2591m18.080s user 2591m17.920s sys 0m0.150s

update: 0x3e2e2292feeb6f4b8019b3e59cd564cb95fff38e0 is 5679780382528065079883438054186157407164239395040 or 5.679780382528065079883438054186157407164239395040e+48

Replies are listed 'Best First'.
Re^6: Pattern enumeration.
by BrowserUk (Patriarch) on Aug 02, 2010 at 07:50 UTC

    I'd already switched to the second version, but not the third. When you posted the third version, I roughly calculated that the time that the second version had already been running more than offset the further 25% saving of switching, so I stuck with it.

    A day or so later I saw your revised runtime estimates. At that point I suspended the process for a while, but then decided to let the code run to completion. It has currently clocked up 100 hours.

    If I am interpreting your numbers correctly--13000 * .6 /60 = 130--it should complete here sometime Wednesday.


    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.
      13000 * .6 /60 = 130

      13000 was the estimation I did based on the timings from the 5 year old computer I use at home.

      I run the 8x8,6 case in a fairly new (and underused) server at work and it took just 43 hours to complete.

        It just now completed after 201 hours. It seems that my AV software was particularly interested in it for some reason, which probably slowed it down.