Tuesday, March 25, 2014

70-461 Training Kit Chapter 1, Lesson 1: Foundation

  • based on set theory & predicate logic
  • Declarative language not procedural.
  • Is 3-valued predicate logic: true, false, unknown
Relational Model
  • not equal Relationship
  • based on 2 mathematics branch: set theory and predicate logic
  • All attribute must has a name
  • All attribute names must be unique
  • No ordering for column or row
  • is a dialect extension of SQL base on ISO standard.
  • Based on multiset theory (may include duplicates).
  • Allows column ordering based on the order that column is create/defined
  •  to use standard syntax such as CAST (standard), CONVERT(T-SQL specific)
  • Your code and knowledge more portable.
A Relation:
  • Is a like a Tables in SQL
  • include a heading and a body
  • Heading of a relation is a set of attributes
  • Body of a relation is set or rows(tuples)
    like rows

A Set:
  • consider as a whole (interact with a set as a whole not row by row, RBAR)
  • no order
  • no duplicate
A Predicate:
  • a proposition either true or false
  • for filter
  • enforce data integrity
  • define data model (???)
An attribute
  • identify a name and a type
Relational thinking
  • Avoid thinking iterative (cursor, loop).
  • Think in set.
When use order by clause
  • The result is considered as a cursor, not relational.
  • Missing
  • NULL doesn't equal blank
  • NULL doesn't equal NULL
  • Need to handle with care.
  • Is a mark for a missing value - not a value itself.
Using correct Terminology

Fields and Records
  • Are physical
  • Field <> column
  • Record <> row
Tables are logical
  • Have logical rows and columns.
Geeks speak: Not NULL value, said NULL mark instead

Credits: Microsoft 70-461 Training Kit

No comments:

Post a Comment