I have a multiple linear regression model with one output value and two input values.
z=Ax+By+C
I would like to plot a graph of residual errors vs instances. Is there any standard tool which I can use. I have the data and to use Openoffice calc, I can calculate SLOPE and INTERCEPT from inbuilt functions but they can be used for a simple linear regression only. HOw can I use calc. here.
Cheers.
For your data, your fitted regression surface is a plane, therefore your residuals are naturally visualised as a perpendicular height field above or below the (x,y) coordinate of each observation point (unless this was obtained using orthogonal distance regression of course but you don't specify this). Thus you really need a surface plot and a software package that will allow you to produce one.
Alternatively, if your data points are arranged as a mesh, you could produce a series of simply regression plots by effectively fixing each of the x (or y) values in your set and producing a residual plot for each x value. Effectively you are splitting your planar mesh residual plot into a series of parallel lines that can be displayed with your tool.
You do not specify whether you have meshed data or if this is acceptable, so as a suggestion I would take a look at Octave as a display tool. This has the capability of producing 3D surface plots and meshes which is really what you need to effectively display a residual surface obtained from this type of analysis. If you have not used Octave before you will have a bit of a learning curve but it is worth a try unless you get an answer that suits you better.
Here is a web-based, interactive tool for plotting regression results in three dimensions.
This 3-D plot works with one dependent variable and two explanatory variables. You can also set the intercept to zero -- i.e., remove the intercept from the regression equation.
The graphics require a WebGL-capable browser, and the most recent versions of all major desktop browsers support WebGL.
