Re: Question about Date & Darwen <OR> operator
From: paul c <toledobythesea_at_oohay.ac>
Date: Sat, 03 Sep 2005 00:07:19 GMT
Message-ID: <XO5Se.357803$5V4.153773_at_pd7tw3no>
>
>
> I note that the above answer *does* have two tuples with x=1, and
> two tuples with y=a. However if you count them up you will see
> that makes for three, not four.
>
> I don't see how one could argue for the tuple (x=2, y=b) to be
> included,
> since (x=2, y=b) doesn't satisy the condition that x=1 or y=a. Put
> another way, if we project the result over x, we get a value that
> wasn't in the x input relation, and if we project over y, we get
> a value that wasn't in the y input relation.
>
>
> Marshall
>
Date: Sat, 03 Sep 2005 00:07:19 GMT
Message-ID: <XO5Se.357803$5V4.153773_at_pd7tw3no>
Marshall Spight wrote:
> paul c wrote:
>
>>Mikito Harakiri wrote: >> >>>Marshall Spight wrote: >>> >>>>As I understand it, the result would be: >>>> >>>>x y >>>>--- >>>>1 a >>>>1 b >>>>2 a >>>> >> >>i believe the answer is not the above, rather there are 4 tuples, two >>with x=1, ie. one for each of all possible values of the y domain and >>two y=a, ie. one for each of all possible values of the x domain.
>
>
> I note that the above answer *does* have two tuples with x=1, and
> two tuples with y=a. However if you count them up you will see
> that makes for three, not four.
>
> I don't see how one could argue for the tuple (x=2, y=b) to be
> included,
> since (x=2, y=b) doesn't satisy the condition that x=1 or y=a. Put
> another way, if we project the result over x, we get a value that
> wasn't in the x input relation, and if we project over y, we get
> a value that wasn't in the y input relation.
>
>
> Marshall
>
you're right; the '2 b' tuple shouldn't be there.
thanks,
paul c
Received on Sat Sep 03 2005 - 02:07:19 CEST