Thank you that is great, can now get down to some work.

Thanks for all your responses, I only just popped out to lunch and came back to a great set of replies.

I know what to do now, they are all good ways to approach the problem, especially the randcomb argument.

Thank you, that is great, I knew maple had the function it's just that i couldn't remember the exact name of it and the help search in maple is something else.

Wow, that is perfect, thank you so very much doug, I can now get down to some real analysis now with my groups. Everyone on this forum has been a great help, truly a great resource.

Having briefly looked at this solution the equivalence classes have just given back all the subsets of whatever size you require and have not split them up into seperate subsets. Sorry Doug for reposting the question, I thought the old post was pretty dead and needed to repost to the forum to revive it.

The suggestion Doug gave at the end wasn't what i wanted but i managed to use the arguments he gave earlier in his response which benefited me to what i wanted to do. Also, thank you John for providing an elegant solution as well very much appreciate the work. One area which i'm trying to do as well is to represent the classes in their sums so the solution is provided in terms of {a[1]+a[2],a[1]+a[3]} and {a[2]+a[3]} I'm trying to analyse which combinations give which lists of values

Thank you once again Doug, everytime i've been in a sticky situation you have managed to clear it up for me, I'll spend a lot more time getting to grip with these tools and tricks. I've been banging my head against a brick wall for a few days now.

Thanks, I use the V in some other cases, I want to leave it alone in this case

That is great thank you, how would i adjust it so the output is in terms of a[1]+a[2] etc instead of the list output

Thank you robert Israel, that's exactly what i wanted, also thank you doug meade for trying to help