What is the primary difference between a partial dependency and a transitive dep
ID: 3769290 • Letter: W
Question
What is the primary difference between a partial dependency and a transitive dependency? A. A partial dependency only reveals part of a functional dependency and a transitive dependency reveals the entire functional dependency B. A partial dependency’s left-hand side of the equation is part of a key, and a transitive dependency’s left-hand side of the equation is not part of a key. C. A transitive dependency has multiple attributes on the left-hand side of the equation, and a partial dependency only has one. D. There is no difference.
Explanation / Answer
A partial dependency is a functional dependency X -> Y that holds for a relation where X is a proper subset of one of the candidate keys of that relation.
Example 1: A relation R(A, B, C) with candidate key AB and FD B -> C.
A transitive dependency is a functional dependency X -> Y that holds for a relation but which can be considered to transitively follow from two other dependencies X -> Z and Z -> Y where Z is not a superkey of the relation.
Example 2: A relation R(A, B, C, D) with candidate key AB and FD BC -> D. In that case we will also have AB -> BC and AB -> D, since AB is a candidate key. Of these then AB -> D is the transitive dependency since it follows from AB -> BC and BC -> D, but BC is not a candidate key.
The relationship between partial dependencies and transitive dependencies is that if there is partial dependency X -> Y then there is actually also a (particular kind of) transitive dependency. This follows since there must be at least one candidate key, say C, and therefore it must hold that C -> X and C -> Y. So then C -> Y will satisfy the definition of transitive dependency since it follows from C -> X and X -> Y and X is not a superkey.
So in some sense a partial dependency is a special case of a transitive dependency. On the other hand there are transitive dependencies that are not caused by partial dependencies as is illustrated in Example 2.