I was going through my lecture slides when I came across this example which I believe is wrong. My lecturer was unable to clarify. I would appreciate it if I could get some clarification.
R = (A, B, C)
Functional Dependencies = (A -> B, B -> A)
The example stated the the highest normal form for the above is 1NF because A-> B forms a partial dependency.
My solution:
AC -> BC (via augmentation axiom)
BC -> AC (via augmentation axiom)
(A,C) and (B,C) are minimal keys and (A, B, C) are prime attributes.
Am I correct to say this:
If (A,C) is the primary key, A -> B is NOT partial FD as RHS is a prime
attribute. B -> A is a non trivial FD as LHS is not a candidate key.
My lecturer's explanation was that if (A, C) is chosen as the primary key, we dont have to treat (B,C) as a key, if so B is not be a prime attribute, hence partial dependency stands