Department of Electrical Engineering and Electronics
ELEC362
Project
Linear regression application
Module ELEC362
Coursework name Project
Component weight 50%
Semester 1
HE Level 5
Lab location personal computers/laptops, university remote computer
Work Individual
*Estimated time to finish 40 hours (coding and testing)
Assessment method Individually
Submission format Online via VITAL
Submission deadline 11.59 pm on the 17th January, 2021
Late submission Standard university penalty applies
Resit opportunity None except for extenuating circumstance
Marking policy Marked and moderated independently
Anonymous marking No (the marker needs to link the code to the report)
Feedback via VITAL GradeMark® / Turnitin Feedback Studio
Expected release of marks date 15 business days from the deadline
Learning outcomes LO2: Using C++ to implement GUI-based software.
LO3: Using online documentation for self-learning.
*Note: This estimate may vary based on the need to debug your application. Make sure
you start working on the project as soon as possible.
Page 2 of 5
The project
Task 1: Design and implement a Qt-based GUI linear regression application. The application reads a
file of data points, given by two columns one for x coordinates and one for y coordinates, then
calculates the best linear fit of the form:
𝑦 = 𝑎𝑥 + 𝑏
Where “a” and “b” are parameters to be determined by the application. You are requested to work
on the GUI only. The regression part of the application will be done by GSL library. Binaries for linking
Qt to the library will be provided for both Windows and Mac OS on VITAL. The documentation of the
linear regression part of the library can be found in this link. You are not asked to study the theory
behind linear regression. All what you are expected to do is to know how to implement the function
from the library into your application, and how to display its return or output to the user.
The application will be used to do linear regression for one file at a time.
In addition to the core functionality mentioned above, the application:
- Should plot the data and the fitted curve on a single graph in the application.
- Should allow the user to choose the thickness and the colour of the plotted curve using 3
different means.
- Should give the user the option to do the fitting with and without a constant term.
- Should be validated, in the report, by using a test dataset as an input to the application
obtaining the fitted line, then comparing the fit to a linear fit by another programme such as
MATLAB or MS EXCEL (No linking to EXCEL or MATLAB is required in the application).
- Should be able to deal with potential run-time errors, which the developers should identify.
Task 2: Modify the application to have a bilinear fit mode, where the fit of the dataset should be
done using a piecewise function of the form:
𝑦 = {
𝑎1𝑥 + 𝑏1 𝑥 < 𝑥0
𝑎2𝑥 + 𝑏2 𝑥 > 𝑥0
The application should calculate (a1,b1,a2,b2) in additional to automatically identify the point x0
which gives the best possible fit. This part if meant to be as a challenge and as a result no hints will
be given on how to do this part. An example of a bilinear fit is shown in figure 1.
It is an expectation from everyone to consult Qt documentation to find any functions or classes that
might be useful for the implementation of this project. Finally, you should make the programme as
professional as possible. Imagine this project as a real-life project given to you.
Approach to be followed
The following steps will help you to have a good design and implementation of the application:
– Start by conducting a background overview of the available programs that for linear
regression and how they are used. This will give you a feeling of what a professional design is
Page 3 of 5
and a better understanding of the core functionality. Make sure the practical notes given in
the lectures ae taken into consideration in your design and implementation.
– Follow the incremental model by breaking the implementation of the application into smaller
parts, where every part is designed, implemented and tested before other parts are
implemented.
– Make sure your application is working perfectly before moving to task 2.
Note for Mac OS users: You can develop the application and write the report based on Mac version
of the app. The application will be tested and marked on a Windows machine. You are strongly
recommended to verify that your app works as it should on a windows machine before submission,
to avoid any potential cross-platform issue, if you have any problems, please let me know.
Figure 1: Example of a bilinear fit for data.
The deliverables
Every submission should consist of the project’s files + brief report (Do NOT submit GSL library’s files
as I will compile your project with local GSL files on my machine). The report must have the student’s
name and ID number. The report does not need an introduction or a conclusion, it must have a cover
page, and includes the following sections:
• How the programme works (explain the idea of the algorithm, include a block diagram).
• User instructions (how should the user use it)
• Testing and verification attempts.
• Overall notes (optional): in this section you can highlight any special features in your design
or implementation of the application. Or you can mention any major issue you faced how
you overcome it. If you left this section empty you will not lose any marks.
• Appendix: headers+ source files (must be in machine readable format).
X0
a2 , b2
here