CPSC 220 Fall 2007
Program 1: Employment Data

Background

"...to encourage and facilitate local government cooperation and state-local cooperation in addressing on a regional basis problems of greater than local significance. The cooperation resulting from this chapter is intended to facilitate the recognition and analysis of regional opportunities and take account of regional influences in planning and implementing public policies and services. The planning district commission shall also promote the orderly and efficient development of the physical, social and economic elements of the district by planning, and encouraging and assisting localities to plan, for the future." (Information from An Introduction to Planning District Commissions and the Roanoke Valley-Allegheny Regional Commission, http://www.rvarc.org/about.htm.)

One such planning commission is the Roanoke Valley-Allegheny Regional Commission (RVARC), located on Luck Avenue in Roanoke. Among its many functions, the RVARC uses employment data collected by the Virginia Employment Commission (VEC) to identify areas of economic growth in the region. The VEC data provides the following information about each employer in the region for each quarter of each year:

• Federal code for the city or county in which the employer is located (FIPS)
• Federal code for the type of company (SIC codes -- here's the list if you're interested)
• Trade name, if different from the legal name
• Legal name
• Mailing address (street, city, state, zip, and extended zip). This could be local or could be for a headquarters or processing center located elsewhere.
• Physical Address (street, city, state, zip, and extended zip). This will be in the region.
• Number of people employed for each month in the quarter

The RVARC faces a number of challenges in processing the data it receives from the VEC. One such challenge is that many companies have multiple locations, and for most purposes the RVARC would like to combine all locations for a single company.

Your assignment is to write a Java program that takes a file containing this data, aggregates employment information for employers with multiple locations, and answers queries about the aggregated data. Requirements for input, output, and program structure are given below.

Input

Your program should prompt the user for the name of the data file to use. You can test your program on file employ99-2.dat, which contains the data described above for the second quarter of 1999. The data for each company is stored in 21 comma-separated fields. Download and open the data file and identify exactly what the fields are -- the first line contains headers for each field that should help. You can assume that any data file used will be in the same format, including the header row.

After prompting for the data file, your program should repeatedly prompt the user for the name of an employer, then display information for that employer as described below. Allow the user to look up as many companies as he or she wishes, with some nice way to exit the program when finished.

Output

• Trade name, if different from legal name
• Legal name
• Type of company (SIC code for now)
• Total number of people employed for each month in the quarter
• Number of physical locations
• Physical address (street, city, state, zip, and extended zip) for each location, and number of people employed at each address. These should sum to your total above.

Program Structure

• LinkedBinarySearchTree -- implements a classic binary search tree. Use the code from the text (you already downloaded this in the jss2 directory), but note that it relies on the find and contains methods from the LinkedBinaryTree class, which are O(N). Rewrite these so that they use the BST properties to produce O(log N) searches (assuming a balanced tree). You may need to complete some of the other methods as well.

• Employer -- stores the information for a single employer. Note that you will need to be able to store multiple addresses and employment data for each address in addition to total employment data. Also, to facilitate searching, you might want to store an additional string that contains the trade name if there is one, otherwise the legal name.

Your program should operate in two phases:

• Build the tree -- Create a BST (initially empty) of Employer objects. Each time you read an employer from the file, check the tree to see if this employer is already there. If not, add it to the tree. If so, do not make a new entry in the tree; simply update the existing entry with the additional employment and address data.

• Respond to user queries -- When the user asks for information about an employer, find that employer in the tree and display the information. No calculations should be required at this time. If the employer is not in the tree, print a helpful message saying so.

Documentation

• Use good variable and method names. This will make life easier for you and for the reader.

• For each method include a header describing the parameters and the return value, if any, and clearly indicating any change the method makes to the object's state (memory, input, output).

• Identify important segments of code so that the reader can easily follow the flow of the program without reading every line of code. For example, you might identify the code that reads information for a single employer, check to see if the employer is already in the tree, and add a new employer or update the information in the existing object. Specifics will depend on how you structure your program, of course. Do not document every line of code!

• Provide additional documentation for particularly complex sections of code. If the reader is likely to think "huh?" when trying to understand what you did, enlighten them!

• Turn in a performance report with your program that describes how you tested your program and what, if any, bugs you found. Unreported bugs will be penalized more heavily than reported bugs.

What to Turn In