COMP0004 Part II Coursework
COMP0004 Coursework Part II
Purpose: To develop an application with a GUI to display data.
Submission: There is a single submission deadline for all the coursework in this module: Noon
25th March 2020. You can complete this part at anytime before the deadline.
Submission information is on the Moodle site.
Marking: This part of the coursework is worth 40% of the module mark.
On Moodle you will find a link to some data files on GitHub in Comma Separated Value (CSV)
format. The files contain synthetic patient data, with each line in the file representing one patient.
The example below is the first two lines representing patients in the file patients100.csv:
3e9fd20e-b81c-4698-
a7c2-0559e10361fa,1979-08-24,,999-46-5746,S99951588,X43863756X,Mrs.,Norah104,Stamm70
4,,Hagenes547,M,white,french,F,Hanover Massachusetts US,678 Trantow Overpass,New
Bedford,Massachusetts,02740
71917638-8142-42da-87f1-
fa0ba642f161,1989-10-21,,999-95-8365,S99995552,X3761822X,Mr.,Eloy929,Dickinson688,,,M,wh
ite,english,M,Dover Massachusetts US,431 Armstrong Forge,Upton,Massachusetts,01568
As you can see, each line is a sequence of strings separated by commas. Each string represents a
piece of information about a patient. These are the fields or columns used to describe a patient,
and are in the same order on every line. You can load a .csv file into a spreadsheet (e.g., Excel) to
more easily view the data. In the data files provided commas are used only as separators between
columns, which means a line can be easily split into a collection of column strings by splitting it at
each comma.
The first line of the patients.csv file does not represent a patient, instead it gives the column name
for each column, also separated by commas:
Inadequate
(0-39)
Failed to demonstrate a basic understanding of the concepts used in the
coursework, or answer the questions. There are fundamental errors, code will not
compile, or nothing of significance has been achieved. (F)
Just
Adequate
(40-49)
Shows a basic understanding, sufficient to achieve a basic pass, but still has
serious shortcomings. Code may compile but doesn’t work properly. Not all
concepts have been understood or all the core questions answered correctly. (D)
Competent
(50-59)
Reasonable understanding but with some deficiencies. The code compiles and
runs, the concepts are largely understood. This is the default for a straightforward,
satisfactory answer. Most core questions answered. (C)
Good
(60-69)
A good understanding, with possibly a few minor issues, but otherwise more than
satisfactory. The code compiles, runs and demonstrates good design practice.
Most expectations have been met. All core questions answered. (B)
Very Good
(70-79)
A very good understanding above the standard expectations, demonstrating a
clear proficiency in programming and design. All core questions plus at least one
challenge question answered. (A)
Excellent
(80-100)
Programming at a level well above expectations, demonstrating expertise in all
aspects. This level is used sparingly only where it is fully justified. All questions
answered. (A+, A++)
of 1 5
COMP0004 Part II Coursework
ID,BIRTHDATE,DEATHDATE,SSN,DRIVERS,PASSPORT,PREFIX,FIRST,LAST,SUFFIX,MAIDEN,
MARITAL,RACE,ETHNICITY,GENDER,BIRTHPLACE,ADDRESS,CITY,STATE,ZIP
The meaning of the names is obvious, for example FIRST and LAST give the first and last name of
each patient (note this is synthetic data so the names have a numerical code attached - just treat
these as part of the name).
If you look through the data you will see that for some patients a column entry may be left blank,
and you see two commas in a row (i.e., Stamm704,,Hagenes547). Beware that the final column
(ZIP) can be left blank, so that there will be a comma followed by newline.
The first column, ID, is a unique id for each patient. The id is effectively the primary key for each
patient and is what you use to uniquely identify each patient. A few other columns such as SSN
and PASSPORT should be unique but are not usually used as primary keys. The rest of the
columns will not contain unique values and you will certainly find that names and addresses
appear multiple times.
There are four patients files on GitHub:
• patients100.csv - 100 patients
• patients1000.csv - 1000 patients
• patients10000.csv - 10000 patients
• patients100000.csv - 100000 patients
These are all in the same format. For development you want to work with the smallest file, but your
code should work with the larger files as well.
In this coursework the aim is to create a general purpose data structure called a DataFrame, that
can be loaded with the patient data (or potentially any other data) and used in a basic dashboard
program with a graphical user interface (GUI). A dashboard is an application that allows you to
view and manipulate data.
A DataFrame has this structure:
This is a collection of named columns where each column holds the data from a single column in
the csv data. For example, the first column would hold the ID for all patients, the second the
BIRTHDATE and so on. A single row across all the columns holds the complete line of csv data for
a single patient, so that row 0 is the data for the first patient and so on. This is the same kind of
layout you would see if you open the .csv file in a spreadsheet application like Excel.
of 2 5
COMP0004 Part II Coursework
Core Questions
For the core questions assume that all the data held in a DataFrame or Column is always of type
String even if there are numerical values or dates. The questions form a specification for the basic
dashboard program, so they don’t have to be answered in order and you may well find it easier to
answer each question in stages as you develop the program.
Q1. Write a class Column, where a column has a name and an ArrayList of rows. Provide methods
getName, getSize, getRowValue, setRowValue, and addRowValue (to add a new row). Determine
what the parameters and return types should be. If you find additional methods are needed when
answering later questions then add them, but only when you need them.
Q2. Write a class DataFrame to hold a collection of Columns. It should have the following methods:
• addColumn
• getColumnNames (a list of names in the same order as they are stored in the frame).
• getRowCount (the number of rows in a column, all columns should have the same number of
rows when the frame is fully loaded with data).
• getValue(columnName, row) to get a row value from a column.
• putValue(columnName, row, value) to put a value into a row in a column.
• addValue(columnName, value) to add a value to a column (at the end).
Again you can add methods if you find you need them when answering the later questions. Also
remember that all the data is in String format.
Q3. Write a class DataLoader that can read a .csv data file and load the data into an empty
DataFrame. The Column names are found as the first row in the .csv data file. It should have a
method that returns a filled DataFrame.
Note that the file input methods will require you to manage IOExceptions that can be thrown (e.g.,
trying to open a file that doesn’t exist). You will have to decide how to manage these. Your program
should not crash or terminate if an exception is thrown!
Q4. The program should have a Swing GUI which should communicate with a class Model to
access the data to display. The GUI should only use the Model to get data and not use or know
about DataFrames or Columns. Class Model provides a wrapper or Facade that hides how data is
stored, providing a set of public methods the GUI can use as needed. While not part of this
coursework the aim of having the model is to allow alternate data storage (e.g., a SQL database)
to be used without having to change any of the code that uses the Model.
Write a class Model that manages a DataFrame internally and provides the methods that your GUI
needs, and uses the DataLoader to load data into a frame.
Q5. Write a Swing GUI that allows the data to be viewed in various ways. For example a basic GUI
might look like this:
of 3 5
COMP0004 Part II Coursework
Here the list of column names is displayed as check boxes that can be selected to determine what
to display in the main panel, and there is a load data file button. You may want to tidy up the
display and provide more (or better) ways of displaying the data!
Q6. Write a Main class that contains the main method and any code to initialise the program, such
as creating the GUI and Model objects. This class should contain the minimum code needed to get
the program running, so will be quite short. The code for building the GUI out of Swing components
should be in the GUI class.

Challenge Questions
Q7. Add the ability to search the data and perform operations such as finding the oldest person,
the number of people living in the same place, and so on. The GUI will need to be extended to give
access to this functionality. Remember that the GUI only displays data it can get from the Model,
so the implementation of search and other functions should be in the Model or other classes, not in
the GUI code. Input from the GUI should be passed as parameters in method calls to the Model.
Q8. The specification in Q1-6 does not state that any Java interfaces should be used (as in
programming to an interface). Identify where interfaces can be introduced and add them, modifying
the rest of the code as needed. Remember interfaces are used to separate specification (a set of
methods) from implementation, allowing different implementation objects to be substituted. For
example, to have DataFrame implementations that store the data in different ways internally. Think
carefully about how and where implementation objects are created, remember the Factory pattern.
Q9. Write a class JSONWriter that given a DataFrame writes out the data to a file in JSON format.
The JSON should be well-formed. You have to write all the code to generate the JSON yourself
and cannot use a JSON library. As you will be doing a lot of String manipulation investigate the
StringBuilder class, as this is much more efficient at building strings than the String class itself. Add
a save to JSON format functionality to your GUI.
Also write a JSONReader that can read data files in JSON format, rather than the .csv format, and
get the data into a DataFrame.
Q10 (hard). The DataFrame for Q1-6 just stores data in String format. Modify the DataFrame and
Column classes so that each column can hold data of a specific type such as Integer, Double or
any other class that represents some kind of data value. You will need to add information to the
classes about the type of values that are stored, make significant use of generics, and modify a lot
of the other code. You should aim to do this without using cast expressions (e.g., forcing a type
conversion in an expression like (int)n) and to use static typing (i.e., type checking done by the
compiler) as much as possible.
of 4 5
COMP0004 Part II Coursework
When a Column stores a specific type of numerical data it enables operations such as Sum,
Average or Standard Deviation to be directly calculated without having to convert Strings into other
types.
Hints: investigate the generic library class Class that represents class types at runtime, the
wildcard operator in generics (e.g., the question mark: List), and the Typesafe Heterogeneous
Container Pattern for Java.
of 5 5