0

I'm trying to figure out the dependencies for the table. However, I've realized that the primary key is composed of a single candidate key which is the Ticket#. I know that every Parking Ticket # will be unique every instance. With that in mind, doesn't that mean this table is already normalized? Other examples in my textbook always had a composite primary key. I've tried going through the process of normalizing anyways.

For my fully functional dependencies I had: Ticket# -> (all other attributes) For my partial dependencies: Ticket# -> (LicNo Date, Code, Fine) For my transitive dependencies: St_ID -> (LName, FName, PhoneNo, StLic) & Code-> (Fine)

Am I doing something wrong? Should I have my StID as part of the primary key with Ticket#?

enter image description here

philipxy
  • 14,867
  • 6
  • 39
  • 83
sangmin park
  • 547
  • 3
  • 7
  • 16
  • 1
    You need to eliminate repeating groups and review the concepts of partial and transitive functional dependencies – reaanb Mar 08 '18 at 19:09
  • Even if I remove the repeating groups, eliminating the null values, wouldn't I still have a primary key with a single attribute? I thought partial dependencies were attributes that were dependent on key-attributes and transitive dependencies are attributes that are dependent on non-prime key attributes? – sangmin park Mar 08 '18 at 19:15
  • 1
    A functional dependency A → B is said to be partial if removal of any subset of A still determines B uniquely. A functional dependency A → C is said to be transitive if there's attributes B such that A → B → C. – reaanb Mar 08 '18 at 19:49
  • 1
    @reaanb ... and not B → A. – philipxy Mar 10 '18 at 07:49
  • Hi. Please [use text, not images/links, for text (including code, tables & ERDs)](https://meta.stackoverflow.com/q/285551/3404097). Use an image only for convenience to supplement text and/or for what cannot be given in text. – philipxy Mar 10 '18 at 07:54
  • You don't mean "a single CK", you mean "a single attribute". There can be multiple CKs & all affect NFs & the FDs make the CKs. Read & use the definitions of the terms you are using. You are not following correct procedures. IMPORTANT: What textbook are you trying to follow? What is your work step by step following it? Because now you are just asking us to rewrite your textbook. Also, explain exactly what you mean by "normalized". That is used for 1NF, which has no standard meaning, *and* for normalization to higher NFs. PS Edit clarifications seamlessly into posts, don't put then in comments. – philipxy Mar 10 '18 at 08:00
  • Eg The definitions of NFs involve *all* CKs. So it's not enough to just know one. [Some textbooks give initial special-case definitions for when there is known to be only 1 CK, which they will refer to as PK.](https://stackoverflow.com/a/49122615/3404097) Eg The notion of PK is not needed in relational theory. Eg *explain that "table"*. Is it a relation formatted in a certain way involving not showing certain values? Something else? How are we supposed to know? Please act on my comment so we can help you. PS Search (my) [so] posts 'user:3404097 every subset superset fd ck'. – philipxy Mar 10 '18 at 21:08

0 Answers0