-1

If I have a table {A, B, C} and the FDs: {A -> B, B -> C} is the table in 2NF ?

Im a little bit confused because of the transitivity axiom which is if A -> B and B -> C then A -> C.

Is C FULL functional dependent from A or is C just normal functional dependent from A ? And in which NF is the table ? 1 or 2 ?

philipxy
  • 14,867
  • 6
  • 39
  • 83
Bobi
  • 11
  • 1
  • Does this answer your question? [Partial Dependency (Databases)](https://stackoverflow.com/questions/25747802/partial-dependency-databases) – philipxy May 07 '22 at 19:36
  • 1
    Please ask 1 (specific researched non-duplicate) question re the 1st place you are stuck/unsure (i a process you give).PS ["1NF" has many meanings.](https://stackoverflow.com/a/40640962/3404097) (All involve replacing some table with parameterized structure by some table(s) with a column per parameter.) So if you ask a question about it, define yours. – philipxy May 07 '22 at 19:37

1 Answers1

0

The relation is in 2NF, but not in 3NF. You are confusing full and partial dependencies with transitive and direct dependencies.

A dependency like A -> C is a transitive dependency when we have that A -> B and B -> C.

On the hand, a relation is not in 2NF if there are partial dependencies: that is if a non-prime attribute depends on a part of a candidate key, that is on a subset of the attributes of a candidate key (so there must be a candidate key that is formed by more than one attribute).

Renzo
  • 26,848
  • 5
  • 49
  • 61
  • They are apparently also confusing Armstrong's transitivity axiom for deriving FDs with the definition of transitive FD. – philipxy May 07 '22 at 20:33
  • That is not a correct definition of transitive FD. Also it is unclear, especially for someone who is lost. Also as usual it looks like they don't understand that they have a cover not a closure & that the relevant definitions talk about all the FDs in a closure. Also by "part" you aren't clear that you mean "part that is not all of". Also "there must be a candidate key that is formed by more than one attribute" is not true, {} can be a determinant violating 2NF when all CKs are single attributes. – philipxy May 07 '22 at 21:31