Re: constraints in algebra instead of calculus

From: paul c <toledobythesea_at_oohay.ac>
Date: Fri, 15 Jun 2007 15:18:37 GMT
Message-ID: <h5yci.30530$xq1.3113@pd7urf1no>

paul c wrote:
> Bob Badour wrote:
>

>> Jan Hidders wrote:
>> ...
>>
>>>
>>> You can simplify that a bit:
>>>
>>> R1 := (R{B}) GROUP {B} AS gB
>>> R2 := (R GROUP {B} AS gB){gB}
>>> constraint: R1=R2

>
> ...
>

>>
>> Hi Jan,
>>
>> That constraint looks like a tautology to me. Can you explain how any 
>> relation with a B attribute could fail the constraint?

>
>
> Butting in, pardon the sloppy notation, isn't this an example?
>
> R:
> A B
> 1 1
> 1 2
> 2 1
>
> R1:
> gB
> {1,2}
>
> R2:
> gB
> {1,2}
> {1}
>
> p

Oops, please ignore that.

p Received on Fri Jun 15 2007 - 10:18:37 CDT