CS14 - At-Home Programming Assignment 4

Project due Thursday, Feb 24 at 8:00 pm

In this project, you will write a program to hash 5 letter words. You will decide on your own hash function. Use seperate chaining. You must use the Standard Template Library list class for your chains. The program will consist of two parts: passing the functionality tests and the success of your hash function. Your grade will be based on the functionality tests. We will then compare the success of your hash function will the other students in the class. Extra credit will be given to a few of the top hashing functions. For us to test the functionality, you need to write at least the following functions in your Hash class. You must use the function prototypes provided. If you change these function prototypes at all, your code will not interface with the test file and your project will not be graded.

void print ( ) - Print the hash table in the following format. Each array location will be printed on a seperate line. The line will start with the number of the array location followed by a : and followed by one tab. You will then print each word stored at that hash array location seperated by a , and a single space. After all words at that hash array location are printed, you will output a return. Here is a sample of what your output should look like
void processFile ( string filename ) - Add all of the words in the file specified as "filename" to the hash table. Here is a sample input file. Use push_back to insert each of the words into the list at the hash location that the words hashes to.
bool search ( string word ) - Search for the string "word" in the hash table. Return true if it is found and false if it is not found
void remove ( string word ) - Remove the string "word" from the hash table
void output ( string filename ) - Output the hash table to the file specified as "filename." Print instructions are the same as listed above.
void printStats ( ) - Print the hash statistics. During program execution, you will need to keep track of some statistics.

Keep a count of the number of collisions when doing inserts. A collision is when a word is being inserted into a non-empty list, which will add 1 to the count (regardless of how many words are already in the list). All other operations should not affect count in any way
Keep track of the length of the longest list
Keep a track of the running average list length (this means that you will update this value after EACH SUCCESSFUL insert and remove operation and not simply find the average list length at the end of the program). So after EACH insert or remove operation, determine the average list length of the current hash table. Then average this with the previous average list length. Your average list length must match mine exactly (you will see a sample below). So, to make this a bit more clear, your average length should start at 0 and each time a successful insert/remove is done update the average length. Here is a formula you can use to update your average length: avgLength = (newAvgListLen + avgLength) / 2.0

The Hash Function

You will need to write your own hash function for the project. You will need to write a unique and original hash function so that you can be eligible for extra credit (you are competing the creativity of the other students). Your hash function must be contained within hash_function.cc. Hash_function.cc already contains a very simple and very poor hash function. You will need to delete the contents of the hash function and replace it with your own hash function. You MUST place your hash function code in the hash function in hash_function.cc and link this file to your project with your makefile. This is extremely important because to test the functionality of your program, we will be replacing your hash function with our own. Actually, we will simply download a fresh copy of hash_function.cc and compile and run your program. If we are unable to do this, you will lose points on your project. Like the main.cc file, DO NOT put any code necessary for correct operation of your program in hash_function.cc besides your hash function.

At the top of this file, you will need to describe your hash function, why it is good, and how you happened upon it.

Defining the hash table size

The size of your hash table should be specified as a defined value. If you look at hash_function.cc you will see that there is a value HASH_TABLE_SIZE that is not defined in this file. For this assignment, you will be learning something new. Using a command line parameter to g++, you can define the value of constants used within your program during compilation. This is accomplished using the -D flag (usage = "-D HASH_TABLE_SIZE=10"). This example sets the hash table size to 10 for your program. This table size is extremely small so I suggest you test your program with table sizes into the thousands. So for this project, the flags to g++ in your makefile will be "-Wall -W -Werror -pedantic -g -D HASH_TABLE_SIZE=X" where X will be replaced with whatever size you want your table to be. Using this method will make it very simple for us to test your programs for different situations.

Provided Files

hash_function.cc - Put your hash function in the function in this file and link this file in with your makefile. DO NOT include your the file by having #include "hash_function.cc" in your hash.cc file. This is very, very, very bad. You MUST link in the hash_function.cc file with your makefile. (Remember to be creative so that you can compete for the extra credit). Like the main.cc file, DO NOT put any code necessary for correct operation of your program in hash_function.cc besides your hash function.
hash.h - Your hash class with the declarations of the functions needed to interface with main. The variables and function prototypes provided are by no means inclusive and you will most likely need to add to this class
main.cc - We will be using this main file to test your program. This means, the first thing that we will do when grading is delete your main.cc and replace it with ours. DO NOT put any code necessary for correct operation of your program in your main.cc file. The output of your program must look EXACTLY like sample.out using the provided hash function in hash_function.cc and a hash table size of 10. This means that you need to do a diff between our sample output and your actual output to determine if your program will pass our testing (do a man on diff to read about it). Be advised, that if your output does not look EXACTLY like our output, you will lose points. This means that diff would output nothing when comparing our sample output to your sample output. In addition to main.cc, we may use additional test cases or test files. If I don't test something in main.cc do not simply assume that you don't need to handle it and that it won't be tested.
sample.out - Sample output obtained by using main.cc and hash_function.cc
test1.txt - The sample input file used for testing main.cc.
hash.out - The output obtained by printing the hash table in main.cc to the output file. Your hash.out must look EXACTLY like this one. Do a diff to make sure.

The hash competition

Extra credit will be available for the top few hash functions (don't ask how much extra credit or how many students will get it.) To be eligible for the hash competition, your hash statistics must match mine EXACTLY during the functionality testing phase (see the stats at the bottom of sample.out and make sure that yours are EXACTLY the same. The reason for this is fairness. We have to make sure that you tally your stats correctly before we can compare your stat with the other students. We will test table sizes of 2004, 10007, and 15659 with dict5.txt and random.txt. Use these files and these table sizes to tune your hash function. Additionally, we will be recording the time that your hash function takes to execute and the time will also be considered when determining the winners. You want to create a good hashing function but if it takes too long to execute, the hashing function is not as useful.

Special note - you must have a unique hashing function to get extra credit. If the top hashing functions are too similar, those students will share a small amount of extra credit. This is to ensure that you do not talk with the other students about your hashing function.

Programming

Be sure to use good programming where possible (use of classes, functions, good parameter passing techniques, operator overloading, templates, error checking). Part of your grade will be based on your actual coding.

What to turn in

You should have turned in what you have thus far on our program at least 6 hours before the due date (by 2pm). Then continue to work on your program and turn in more current versions as you get them working. Ideally, you should have your project done well before the due date.

You must turn in all .cc and .h files and your makefile. In addition, you must follow any guidelines and include any information that is stated on the main page about at-home programming assignments.

DO NOT, I repeat, DO NOT turn in any files that are not required for the final compilation of your assignment. Do not turn in your temporary .cc files or your saved/olded revisions of your .cc files. If you turn in files that aren't required for your assignment, it hinders the grading processes. You will be docked points if you turn in extraineous files.

A reminder about collaboration on home programming assignments

Please remember, at-home programming assignments are not lab assignments and you may not team-code with your lab partner or any other individual. Limited collaboration may be acceptable, but programs must represent YOUR OWN original work. Sharing code or team-coding are strictly not allowed. Copying code from ANY source (any book, current or past students, past solutions, web, etc) is strictly not allowed even with citation. Collaboration may consist of discussing the general approach to solving the problem, but should not involve communicating in code or even pseudo-code. Students may help others find bugs. Your code MUST be unique -- the odds of randomly producing similar code is very low. Computing, like surgery or driving a car or playing golf, can only be learned by doing it yourself!