Regression

Regression allows for extrapolation or prediction. Regression “fits” given data to a curve, and then uses the equation of the curve to predict points both within the boundaries and beyond the given boundaries.  It’s interpolation on steroids, but be careful - you need to have a feel for your data!.

To get the code, please go to the page dealing with:

Linear Regression

Fits to f(x)=a+b*x

Geometric Regression

Fits to f(x)=a*x^b

Exponential Regression

Fits to f(x)=a*e^(b*x)

Nth Order Regression

Fits to f(x)=c+a1*x+a2*x^2+...+aN*x^N

Regression of an Arbitrary Function in One Variable

Fits to a user supplied form.

Polynomial Regression in Two Variables

An Nth order regression for a function of two independent  variables.

Multiple Linear Regression

A least squares linear regression for multiple independent variables.

 

A GUI program that combines four of the above methods with plotting:

Liberty Basic 3 - source code

Liberty Basic 3 - executable (wrapped with run time engine)

Links:

http://www.math.csusb.edu/faculty/stanton/m262/regress/regress.html

Nice linear regression link.