100 CSE W.A. Burkhard
Programming Assignment Two Due: midnight Friday February 17, 2012 Separate Chaining Hashing
Hashing provides an excellent table implementation technique with expected access times independent of the table size. We implement a polymorphic instance of the table data type using separate chaining hashing together with the move-to-the-front heuristic to determine it's performance experimentally. Your assignment is to create the table implementation in C++. Table Data Type Here is a portion of the Table header describing the functionality you must provide within your implementation.
1: 2: 3: 4: 5: 6: 7: 8: 9: 10: 11: 12: 13: 14: 15: 16: 17:
typedef struct { double successful[2] , unsuccessful[2] ; } Perform ; template <class DATA> class Table { private : // your data structure implementation choices here public : explicit Table ( unsigned int size = 5 ) ; void clear ( ) ; bool insert ( DATA & data ) ; bool remove ( DATA & data ) ; bool access ( DATA & data ) const ; Perform perform ( ) const ; list<DATA> & getOne ( unsigned int i ) ; } ;
The explicit Table constructor, with the default parameter, creates an empty table for DATA objects. The explicit "modifier" avoids type conversion without explicit type casting. The default parameter assures a table of size five will be built unless another size is explicitly requested. The clear method resets the table data structure to its initial state so it contains no DATA objects. The destructor is called for each object. The insert method includes the DATA object within the table using separate chaining hashing and returns true if successful. The method returns false if a duplicate is found. The object is included at the "back" of the appropriate list. The access method locates the object within the table and sets data to this value as well as returning true. Otherwise, false is returned. The data parameter initially contains only the key for the desired record. This method implements the move-to-the-front scheme; the STL erase function is a suggestion in this context. The delete method removes the object from the table and sets data to this value as well as returning true. Otherwise, false is returned. The data parameter initially contains only the key for the to be removed record. The perform method determines the average successful and unsuccessful search lengths for the table and returns them via a Perform structure. The average successful search length is determined for all the records within the table. The average unsuccessful search length is determined from the lengths of the chains. These values are stored in perform successful[0] and unsuccessful[0]. The expected successful and unsuccessful search lengths are calculated and returned in successful[1] and unsuccessful[1] respectively. The getOne method returns a list<DATA> reference providing access to the ith list. This is not a "normal" table method it but is useful in the course. The private portion within the class Table declaration provides our implementation hiding/security mechanism.
version 2
Main Programs There are four main programs main0, main1, main2, and main3 for your use. Program main0 demonstrates the example from page 40 of the notes. The program main1 allows you to create performance statistics for separate chaining. The main2 program tries to insert a duplicate record as well as determining where a record is stored. Finally program main3 stores more realistic records with a useful hash function. There is an optional makefile for these programs; the makefile assumes your table.hpp and table.cpp files are in the same directory. Hints 1. Begin by trying to build very small tables with only a record or two. You can modify the main0.cpp program to build such tables. 2. Your class Table data structure must be "private." What to Turn in You are to submit the following files: table.cpp, table.hpp, README as a tar file affectionately referred to as a "tar-ball." Your table.hpp is to differ from the one in the homework folder only within the private section.
-2-
Sample Output list 0: 1: list 0: 1: list 0: 0 12 18 1 16 22 2 20
pig ape dog yak cow average usl 1.666667
average ssl 1.400000 list 0: 1: list 0: 1: list 0: list 0: list 0: 1: list 0: 0 12 18 1 22 16 2 20 0 12 1 22 16 2 20
pig ape yak dog cow
pig yak dog cow average usl 1.333333 Figure 2. main0 output
average ssl 1.250000
-3-
load factor 0.000000 0.100000 0.200000 ... 9.700000 9.800000 9.900000 10.000000
average ssl 1.000000 1.000000 1.045000 5.811753 5.835000 5.912525 5.938050
expected ssl 1.000000 1.000000 1.050000 5.800000 5.850000 5.900000 5.950000
average usl 1.000000 1.000000 1.009000 9.700000 9.800000 9.900500 10.000000
expected usl 1.000000 1.000000 1.010000 9.700036 9.800033 9.900030 10.000027
Separate Chaining Separate Chaining Separate Chaining Separate Separate Separate Separate Chaining Chaining Chaining Chaining
Figure 3. main1 output:
created via main1 10
-4-
list list 0: 1: 2: list 0: 1: 2: 3: list 0:
0 1 2345 6781 3589 2 1234 2342 234 9982 3 235
record 6781 is present record 7766 is not present record 98 is not present duplicate record attempted: 6781 average ssl is 2.125000 average usl is 2.250000 expected ssl is 1.875000 expected usl is 2.100113 list list 0: 1: 2: list 0: 1: 2: 3: list 0: list list list list 0 1 6781 2345 3589 2 1234 2342 234 9982 3 235 0 1 2 3 Figure 4. main2 output
-5-
list 0 0: name: list 1 0: name: 1: name: 2: name: list 2 0: name: list 3 0: name: 1: name: 2: name: 3: name: 4: name: average ssl expected ssl
alex kronkov beth smith darrell cruz willie yuu herb jones alan knox patti bliss jehan roma diane lake bev tuxin is 2.300000 is 2.125000
zip: 56789 zip: 92014 zip: 23455 zip: 12345 zip: 92093 zip: 78901 zip: 16428 zip: 34567 zip: 16428 zip: 92014 average usl is 2.500000 expected usl is 2.556314 Figure 5. main3 output
-6-