From: Lyndon Samson <lyndon.samson@...>
Date: 2005-10-19T08:37:43+09:00
Subject: Re: AND THE WINNER IS... [WAS] Re: How to get non-unique elements from an array?

------=_Part_21427_18098516.1129678659880
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On 10/19/05, pauldacus@gmail.com <pauldacus@gmail.com> wrote:
>
> Yes... my solution seems to suck in that it doesn't work & is slow. I
> believe this makes me Management Material!
>
> But hopefully I can redeem myself with this humble solution (does seem
> to work), which borrows heavily from my predecessors:
>
> one =3D {}; two =3D {}; a.each{|i| one[i] ? two[i]=3D1 : one[i]=3D1}; two=
.keys


Basically my ( fast :-) )solution using Hashs rather than a Bit fields...

bn =3D bn2 =3D 0
a.each {|el|
bn2 |=3D (1<<el) if bn[el] =3D=3D 1
bn |=3D (1<<el)
}


It uses the much acclaimed one-pass method, but skips the 'uniq'
> method, which seems like it's got to be expensive on a big array. No
> sorting or counting, just a straight up "If this is not the first time
> I've seen this thing, then it's the second, otherwise... it's the
> first". I'd benchmark, but a copy/paste on previous code blows up. I
> blame Windows.
>
>
>


--
Into RFID? www.rfidnewsupdate.com <http://www.rfidnewsupdate.com> Simple,
fast, news.

------=_Part_21427_18098516.1129678659880--