440 likes | 603 Views
Chapter 8: Relational Database Design. First Normal Form Functional Dependencies Decomposition Boyce-Codd Normal Form Third Normal Form. First Normal Form. R is in first normal form if the domains of all attributes of R are atomic
E N D
Chapter 8: Relational Database Design • First Normal Form • Functional Dependencies • Decomposition • Boyce-Codd Normal Form • Third Normal Form
First Normal Form • R is in first normal form if the domains of all attributes of R are atomic • In object relational/object oriented databases, attributes can be composite or multi-valued • But in relational databases, composite attributes will need to be flatten out and multi-valued attributes need to be represented by another relation
Pitfalls in Relational Database Design • We can create tables to represent an ER design in many different ways • Combine attributes differently to create tables • Why do we choose some ways over the others? • Redundancy • Inability to represent certain information • E.g. relationships among attributes
Example • Consider the relation schema:Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • Redundancy, why? • Inability to represent certain information, why? • Cannot store information about a branch if no loans exist • Can use null values, but they are difficult to handle.
Why Redundancy Is Bad? • Wastes space • Complicates updating, introducing possibility of inconsistency of assets value • We know why inability to represent certain information is bad.
Decomposition • Decompose the relation schema Lending-schema into: Branch-schema = (branch-name, branch-city,assets) Loan-info-schema = (customer-name,loan-number, branch-name, amount)
Lossless-join decomposition • All attributes of an original schema (R) must appear in the decomposition (R1, R2): R = R1 R2 • For all possible relations r on schema R r = R1 (r) R2 (r)
Non Lossless-Join Decomposition • Decomposition of R = (A, B) • R1 = (A) R2 = (B) A B A B 1 2 1 1 2 B(r) A(r) r A B A (r) B (r) We do not loss any tuple but we lose the relationship between A and B 1 2 1 2
Relational DB Design Process • Decide whether a particular relation R is in “good” form. • In the case that a relation R is not in “good” form, decompose it into a set of relations {R1, R2, ..., Rn} such that • each relation is in good form • the decomposition is a lossless-join decomposition • Our theory is based on: • functional dependencies • multivalued dependencies
Functional Dependencies • Constraints on the set of legal relations. • Require that the value for a certain set of attributes determines uniquely the value for another set of attributes. • A functional dependency is a generalization of the notion of a key.
Functional Dependencies • Let R be a relation schema R and R • The functional dependency holds onR if and only if for any legal relations r(R), whenever any two tuples t1and t2 of r agree on the attributes , they also agree on the attributes . That is, t1[] = t2 [] t1[ ] = t2 [ ]
Example • Example: Consider r(A,B) with the following instance of r. • On this instance, AB does NOT hold, but BA does hold. • Note: function dependency needs to hold for any possible instance of a relation. • 4 • 1 5 • 3 7
More Example • Consider the schema: Loan-info-schema = (customer-name,loan-number, branch-name, amount). We expect this set of functional dependencies to hold: loan-numberamount loan-number branch-name but would not expect the following to hold: loan-number customer-name (why?)
Keys Defined by Functional Dependencies • K is a superkey for relation schema R if and only if K R • K is a candidate key for R if and only if • K R, and • for no K, R
Example Keys • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • Superkeys? • Candidate keys?
Functional Dependencies • A functional dependency is trivial if it is satisfied by all instances of a relation • E.g. • customer-name, loan-number customer-name • customer-name customer-name • In general, is trivial if
Closure of Attribute Sets • Given a set of attributes a, define the closureof aunderF (denoted by a+) as the set of attributes that are functionally determined by a under F:ais in F+ a+
Algorithm to compute a+ • The closure of a under F result := a;while (changes to result) do for each in F do begin if result • then result := result end
Example of Attribute Set Closure • R = (A, B, C, G, H, I) • F = {A BA C CG HCG IB H} • (AG)+ 1. result = AG 2. result = ABCG (A C and A B) 3. result = ABCGH (CG H and CG AGBC) 4. result = ABCGHI (CG I and CG AGBCH)
Example Closures • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • Closure of attribute set (loan-number)? • Closure of attribute set (customer-name, loan-number)? • Closure of attribute set (branch-name)? • Given: • loan-number branch-name, amount • branch-name branch-city, assets • customer-name,loan-number Lending-schema
Testing for superkey and candidate key Testing functional dependencies Computing closure of F Uses of Attribute Closure
Testing for Keys • To test if is a superkey, we compute +, and check if +contains all attributes of R. • To test if is a candidate key, first make sure is a superkey. Then make sure no subset of is a superkey • In the previous example, is AG a candidate key? • Is AG a super key? • Does AG R? == Is (AG)+ R • Is any subset of AG a superkey? • Does AR? == Is (A)+ R • Does GR? == Is (G)+ R
Example: Testing for keys • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • Is (loan-number) a superkey/candidate key? • Is (customer-name, loan-number) a superkey/candidate key? • Is (branch-name) a superkey/candidate key? • Is (customer-name, loan-number,branch-name) a superkey/candidate key? • Given: • loan-number branch-name,amount • branch-name branch-city, assets • customer-name,loan-number Lending-schema
Testing Functional Dependencies • To check if a functional dependency holds (or, in other words, is in F+), just check if +.
Example: Testing for functional dependencies • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • Does loan-number branch-city, assets hold? • Does customer-name, loan-number, amount amount hold? • Given: • loan-number branch-name,amount • branch-name branch-city, assets • customer-name,loan-number Lending-schema
Computing Closure of F • But what is a closure of F? • Given a set F set of functional dependencies, there are certain other functional dependencies that are logically implied by F. • E.g. If A B and B C, then we can infer that A C • The set of all functional dependencies logically implied by F is the closure of F. • We denote the closure of F by F+.
Armstrong’s Axioms • Sound and Complete rules: • if , then (reflexivity) • if , then (augmentation) • if , and , then (transitivity) • These rules are • sound (generate only functional dependencies that actually hold) and • complete (generate all functional dependencies that hold).
Example • R = (A, B, C, G, H, I)F = { A B, A C, CG H, CG I, B H } • some members of F+ • A H • by transitivity from A B and B H • AG I • by augmenting A C with G, to get AG CG and then transitivity with CG I • CG HI • from CG H and CG I : “union rule” can be inferred from • definition of functional dependencies, or • Augmentation of CG I to infer CG CGI, augmentation ofCG H to inferCGI HI, and then transitivity
Closure of Functional Dependencies • Derived rules from Armstrong’s axioms • If holds and holds, then holds (union) • If holds, then holds and holds (decomposition) • If holds and holds, then holds (pseudotransitivity)
When we compute the closure of attribute set , we already implicitly used Armstrong’s axioms • Let us just use the closure of attribute set to calculate the closure of functional dependency set
Now Compute Closure of F • For each R, we find the closure +, and for each S +, we output a functional dependency S.
Example: Closure of F • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • How to compute the closure of F (F is given below)? • Given F: • loan-number branch-name,amount • branch-name branch-city, assets • customer-name,loan-number Lending-schema
We just talked about the maximal set of functional dependencies that can be derived from a functional dependent set F • Closure of F • What about the minimal set of functional dependencies that is equivalent functional dependent set F • Canonical cover of F
Canonical Cover • Intuitively, a canonical cover of F is a “minimal” set of functional dependencies equivalent to F, having no redundant dependencies or redundant parts of dependencies • What does “equivalent” mean? • What does “minimal” mean?
Equivalent Means: • Give two functional dependency sets F and F’, they are equivalent if and only if: • F logically implies all functional dependencies in F’ • F’ logically implies all functional dependencies in F • We use to represent logically imply and to represent equivalent • How to test if F F’? • For each F’, test if it holds in F • You may need to calculate a+ in F
Minimal Means: • No redundant functional dependencies! • Sets of functional dependencies may have redundant dependencies that can be inferred from the others • Eg: A C is redundant in: {A B, B C, A C} • Parts of a functional dependency may be redundant • E.g. on RHS: {A B, B C, A CD} can be simplified to {A B, B C, A D} • E.g. on LHS: {A B, B C, AC D} can be simplified to {A B, B C, A D}
Extraneous Attributes • For any in F • Attribute A is extraneous ifF (F – {}) {( – A) }. • Attribute A is extraneous if F (F – {}) { (– A)} • Use attribute closures to check equivalence
Examples • Given F = {AC, ABC } • B is extraneous in AB C because {AC, AB C} is equivalent to {AC, AC } = {AC} • Given F = {AC, ABCD} • C is extraneous in ABCD because {AC, ABCD} is equivalent to {AC, ABD}
Canonical Cover • A canonical coverfor F is a set of dependencies Fc such that • F logically implies all dependencies in Fc, and • Fclogically implies all dependencies in F, and • No functional dependency in Fccontains an extraneous attribute, and • Each left side of functional dependency in Fcis unique.
Compute a canonical cover for F repeatUse the union rule to replace any dependencies in F11 and 12 with 112 Find a functional dependency with an extraneous attribute either in or in If an extraneous attribute is found, delete it from until F does not change • Note: Union rule may become applicable after some extraneous attributes have been deleted, so it has to be re-applied
Example • R = (A, B, C)F = {A BC, B C, A B,ABC} • Combine A BC and A B into A BC • Set is now {A BC, B C, ABC} • Is B extraneous in ABC • {A BC, B C, ABC} ? {A C, B C, ABC} NO • Is C extraneous in ABC • {A BC, B C, ABC} ? {A B, B C, ABC} Yes • Set is now {A B, B C, ABC} • Is A is extraneous in ABC • {A B, B C, ABC} ? {A B, B C, BC} Yes • Set is now {A B, B C} • The canonical cover is: {A B, B C}
Example: Canonical Cover of F • Lending-schema = (branch-name, branch-city, assets, customer-name,loan-number, amount) • How to compute the Canonical Cover of F (F is given below)? • Given F: • loan-number branch-name,branch-city,assets,amount • branch-name branch-city, assets • customer-name,loan-number Lending-schema