Sorry to bug everyone with a lame question like this, but does anyone know how to make the chart function give me a quadratic best-fit line? When I right-click the data points and ask it to give me regression line, it only gives me a linear regression. I can't figure out how to make it fit a quadratic relationship.

Can it even do that? I don't even think Excel had that feature.

Oh, and, I'm using OpenOffice 2, not 3. That might matter.

Well I'm running Openoffice 2.3 - where there are just linear, log, exponential, and power. So no obvious quadratic or higher polynomials built in.

This link may be of interest - a series of openoffice worksheets which you can use to perform quadratic fits (among other things). (Note - I needed to rename these downloads from .odt to .ods to load them in Open-Excel).

http://terpconnect.umd.edu/~toh/mode...tionCurve.html

A peripheral comment - If you've got many of these to do it quickly becomes annoying using Excel or OO. You might consider something like python which is a lot easier once you get up and running. For example there is a 'polyfit' module in numpy which is pretty flexible.

http://www.scipy.org/Numpy_Example_List_With_Doc (go down to polyfit)

I think there is also a polyfit in the matplotlib library - not sure if they are the same.

Hope this helps

Steve

You can do it more easily with gnumeric, it has polynomial regression in the graph options.

Thanks for the advice. It's not as though I prefer a spreadsheet to do my graphing for me. I just had to turn in a spreadsheet with a graph for a project recently, and I'd hate to be unable to do a project because I don't use M$ apps.

I hate to deviate from the topic of my own post, but is there anything that Gnumeric can't do that OpenOffice can? I've been looking at it recently, and on the surface, I don't see much of a difference.

Not that anyone really knows the difference until they need something that isn't there. XD

Gnumeric can usually do everything Calc can and more.

Cool. Thanks for all the help, folks.