Relational DBMS.pptVIP

Relational DBMS Motivating Questions From last time: Is MS Excel a DBMS? What makes some databases relational, while others are not? Why do so many people use them? RDBMS: Basic Principles Based on a variant of n-ary relations Relations are the main data structure A relational database consists of 1 or more relations Connections and relationships represented by values of data, not by physical references Reference by value Reference by location Natural representation as tables Relations: Properties Values of each column are homogeneous(belong to the same domain) Rows are different with respect to one another R is a set of tuples (not a bag) The order of columns is irrelevant, they are identified by name, not by position The order of rows is irrelevant, they are identified by content, not by position Relations: Definitions and Properties Some loose definitions: “attribute” corresponds to a named table column; “tuple” corresponds to a table row; “values” correspond to table cells each attribute has a domain set from which its values come from Schemas and Instances Relation schema: A relation name and a set of distinct attribute names: R(A1 A2 ….An) Database schema: A set of relation schema with distinct relation names: R Relation instance defined over schema R (X): finite set r of tuples over XDatabase instance defined over schema R:R = (R1 (X1), R2 (X2), ….Rn(Xn) ) = a set of relations (r1, r2, …., rn) Primary Keys A key is a set of columns that together makes each row distinct No two rows will have the same value of this key (implies that no two rows are allowed to be identical) A primary key is usually included as a single column (e.g., ID) Every relation has a key Foreign Keys in Relations: Example Semantic Integrity Not all finite set of tuples are acceptable : even if tuples have the right degree and attribute values: Data Definition Language (DDL) Used to encode database structure, including: Table Definition (Name, Field Names, Domains,Keys)