CSc 430 - Software Engineering/Advanced GUI Programming
Lab #4 - Prof. Domanski

Download Sample Data for this lab

click on the image below to try it out

Objective To create a generic line graph  with the following required properties:

• read data from a file that a user chooses (e.g. 430DATA.csv)
• allow the user to choose the starting and ending values of the X axis
• allow the user to choose 1 (or more) Y values to graph
• graph the (X,Y) coordinates selected, and appropriately scale the X and Y axes
• compute the linear regression equation for each set of Y's
• graph the linear regression line on the same graph and display R² for each Y selected
• allow users to highlight a new starting and  ending X value from the displayed graph by using their mouse
• redisplay all graphs using the new interval (a zoom function)

The BASIC file structure

Regression Basics Straight Line formula ® Y = M*X + B

Regression: given a set of (x,y)’s, compute approximate values of M and B, so that if given a new value of X, we could estimate what the corresponding Y value would be.

Formulas: where x' and y' are the average values for x and y respectively, and n is the number of points you have --

• Ssx = å x² - n ( x' )²  Ssx = Sum of Squares of X
• Ssy = å y² - n ( y' )²  Ssy = Sum of Squares of Y
• Sxy = å x y - n x' y'    Sxy = Sum of x*y pairs

• M = Sxy / Ssx           M = Slope of line
• B = y' - M * x'           B = Y-Intercept

• Sse = å y² - Bå y - M å x y    Sse = Sum of Squares of errors
• R² = (Ssy - Sse ) / Ssy            R² is a "goodness of fit " estimator: 1 is a great fit, 0 is terrible)

Important Stuff:   

• Working code is due late April (3½ weeks after initial posting)
• Documentation/Help must be included in the working code
• Test Plan notes must be included in the documentation

Last edited by DrB on Wednesday, February 18, 2009