RDBs

Schemas

• Schema is structure of relatable DBs
  • Ex: dept(name, budget, size)
    • name represents the primary key

Keys

• Keys in a given relation are not duplicated (are unique)
• Keys are composed in a hierarchy, where each is more specific than the last
  • Superkeys, candidate keys, primary keys
• Superkeys:
  • Most general
  • Can contain extra info that isn't necessarily needed
  • Ex: if we used (name, size) to reference a dept
• Candidate keys:
  • No extra columns
  • Tables can have multiple candidate keys
  • Ex:
    • If a student relation has both Student_ID and SSN, both are unique to the student
    • Both could be used as reference
• Primary keys:
  • One per table
  • Unique
  • Chosen by DB designer as principal means of identifying a record

Notation

• α typically represents a set of attributes
• When a set of attributes is a superkey, it's denoted by K
• Relation names are lowercase
• r(R) refers to a relation r with schema R
  • r, the relation, is the actual data
  • R, the schema, represents the structure
  • Sometimes, when it's clear an instance is being discussed, the relation name can be used (i.e., r)