卡耐基ssd7_ex6

--ZhangRui 2009303451 RJ010905
1.
List the PRIMARY KEY£º
TrackingNum£ûPK£ý

2.
List all the FDs:
The following FDs involve single attributes
FD0: {TrackingNum} -> {EmpID, EmpName, OrderNo, ShipToAddr, ShippedDate}
FD1: {EmpID} ->{EmpName}
FD2: {OrderNo} ->{ShipToAddr}
In the following FDs, the left-hand-side X involves more than one attribute
FD3: {EmpID,OrderNo} -> {EmpName, ShipToAddr, ShippedDate,TrackingNum}

3.List all the update anomalies and provide an example of each
An insertion anomaly takes place, for example,  when the handlers trying to an EmpID and the EmpName belonging to the new employee who does not have an OrderID.
A deletion anomaly occurs when a deletion leads to an unintended drop of data.For example, information about a employee(like Joe) is lost,when his or her shipment is deleted.
A modification anomaly occurs because some attributes which is repeated.For example,when the "EmpID" is modified in one row ,it is possible for the other row that is not be modified.

4.What normal form is the relation in?
The relation is in an unnormalized form. it does not relate to the 1NF Because the attribute "ShipToAddr" and "ShippedDate" are composite attributes.

5.Apply normalization to it incrementally, bringing the relation to 3NF.
(1) from unnormalized form to 1NF modiify the composite attributes("ShipToAddr" and "ShippedDate") into character strings,if "ShipToAddr" ,"ShippedDate" are character strings; then all attributes are simple atomic and hence SHIPMENTis in 1NF.
(2)from 1NF to 2NF
    EMPLOYEE(EmpID,EmpName)
               EmpID{PK}
               FD: EmpID->{EmpName}
    ORDER(OrderNo, ShipToAddr)
     OrderNo{PK}