From: "Eric J. Whitney" Newsgroups: comp.os.msdos.djgpp Subject: Re: Large arrays and time Date: Wed, 07 Apr 1999 14:34:24 +1000 Organization: The University of Sydney, Australia Lines: 34 Message-ID: <370AE050.29C67D77@aero.usyd.edu.au> References: <7eef5v$r5s AT journal DOT concentric DOT net> NNTP-Posting-Host: pc-114.inthouse.usyd.edu.au Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: metro.ucc.usyd.edu.au 923459790 26106 172.16.79.114 (7 Apr 1999 04:36:30 GMT) X-Complaints-To: usenet AT news DOT usyd DOT edu DOT au NNTP-Posting-Date: 7 Apr 1999 04:36:30 GMT X-Mailer: Mozilla 4.03 [en] (Win95; I) To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Reply-To: djgpp AT delorie DOT com General in GA work you only have to store the current generation and a temporary so that you can generate the next generation. You will find the time is not prohibitive... there just aren't many calculations that need to be done. Once you have finished with a generation, write a summay to a file and forget about it. Use a dynamically created array to hold the individuals. If you are using C++, you can wrap this up in a "population" class. And no, there is no "easier" way to do this... you need a view of the entire population each generation. Hope this helps. Paradox wrote: > I'm doing some reasearch in Genetric Algorithms and I wasn't sure the best > way to store the information about each generation. The best I could come up > with was a struct, and the variable type of the struct would be an immense > array, like a[1000]; The problem with this is the long time it would take > for me to search through the array to find a data "offspring" and relay it's > information to the screen. This will also pose problems for "dying" and > "birth". Is there an algorithm to make this easier, or another way to do > this without arrays. > > Paradox -- --------------------------------- Eric "The Wingnut" Whitney ---------------------------------