I'm trying to do a polynomial curve fit (order 2) for the data below, but the
equation that the curve fit provides is way off. Any ideas why this isn't
working? It works for a similar set of data...

x y
10600 250
10600 250
10250 300
9750 350
9250 400
8625 450
8000 500
6750 550

On Wed, 20 Aug 2008 16:38:09 -0700, rcc wrote:

I'm trying to do a polynomial curve fit (order 2) for the data below, but the
equation that the curve fit provides is way off. Any ideas why this isn't
working? It works for a similar set of data...

x y
10600 250
10600 250
10250 300
9750 350
9250 400
8625 450
8000 500
6750 550

You don't give us any clue as to how your are doing the fit?

Are you using a formula? If so, please post the formula.

Are you graphing it and using the tool within the graph? If so, you probably
need to format the formula to show more decimal places.
--ron

rcc -

(1) Use an XY (Scatter) chart type, not a Line chart type (which treats the
X values as 1,2,3,... for the curve fit).

(2) Be sure all values are numeric, not text.

(3) I get the following results using Trendline or worksheet function
LINEST:

Y = -0.000013147702647*x^2 +
0.1503983217785*x +
134.511365283

- Mike Middleton
http://www.DecisionToolworks.com
Decision Analysis Add-ins for Excel


"rcc" wrote in message
...
I'm trying to do a polynomial curve fit (order 2) for the data below, but
the
equation that the curve fit provides is way off. Any ideas why this isn't
working? It works for a similar set of data...

x y
10600 250
10600 250
10250 300
9750 350
9250 400
8625 450
8000 500
6750 550

Since your data does not have numeric issues in LINEST from any version of
Excel, I presume that you are using the Chart trendline option. Make sure
that your chart is an "XY (Scatter)" chart, since a trendline on a "Line"
chart would give you nonsense.

If you are using an "XY (Scatter)" chart, the usual problem is to try to
calculate with the excessively rounded values of the default display. Right
click on the displayed equation and format to display in scientific notation
with 14 decimal places.

Jerry

"rcc" wrote:

I'm trying to do a polynomial curve fit (order 2) for the data below, but the
equation that the curve fit provides is way off. Any ideas why this isn't
working? It works for a similar set of data...

x y
10600 250
10600 250
10250 300
9750 350
9250 400
8625 450
8000 500
6750 550

but the equation that the curve fit provides is way off.

Hi. You didn't provide your equation. As a side note, another program gave a caution message...
"The abscissae are not distinct."

Re-looking at your data, is this a typo, or did you intend to have an x value with the possibility of having two different y values?


10600 250
10600 250


Again, it was just a caution message.
--
Dana DeLouis


"rcc" wrote in message ...

I'm trying to do a polynomial curve fit (order 2) for the data below, but the
equation that the curve fit provides is way off. Any ideas why this isn't
working? It works for a similar set of data...

x y
10600 250
10600 250
10250 300
9750 350
9250 400
8625 450
8000 500
6750 550