Re: Relation Schemata vs. Relation Variables

paul c wrote:
>
> There must be a subtlety here that eludes me.

I think there is a subtlety here that eludes you.

> If a candidate key, k, of
> a relation has a value of 1 in some tuple and a tuple in that relation
> has a value of {k 1, x 2} then I would say that the value k = 1
> certainly identifies that tuple.

Let's look at it a different way. Let's specify the candidate key as a functional dependency constraint.

Given a relation R with two attributes, k and v, where k -> v. (Aka, {k} is a candidate key.)

Expressing this as a constraint:

forall k, v, k', v' in R: k = k' => v = v'

"For all pairs of tuples (k, v) and (k', v') in R, if k equals k' then v = v'."

So if we are restricting ourselves to only looking at one tuple (k, v), what does that mean? Well, it doesn't mean anything, because what it's saying depends on there being two tuples to compare. If we're talking about a relation with only one tuple, then the pair (k,v) and the pair (k', v') are the same pair, so again it doesn't say anything.

Marshall Received on Thu Aug 24 2006 - 20:47:37 CDT