fitting a curve to charted data

[Dkcorey]

I have a small data set that I have charted and then fitted a trendline (2nd order polynomial).

I then displayed the formula for use in a workbook project. Problem is, the results from the formula differ from the actual data

Here's the data:
The first column are the x values, second column are the y values and the third column are the results given by the formula shown below. The r squared value is .9994

2500 72.444 77.55
2300 67.444 71.89
2100 65.111 68.63
1900 64.778 67.77

y = 3E-05x2 - 0.1157x + 179.3

Thanks for any help!

David

[MrShorty]

The usual mistake that people make in cases like this is to assume that those coefficients are exact. For example, is a exactly 3.00000000000000e-5, or is it 2.5100000e-5 or is 3.4800000e-5. Format the trendline to show more significant figures so you know you are using the correct values for the coefficients.

If you do a lot of curve fitting, you might learn how to generate the coefficients directly in the spreadsheet using the LINEST function. Then you'll avoid the problems that can come from transferring the coefficients from the chart to the spreadsheet.

[Dkcorey]

I adjusted the coefficients by trial and error and got a good result. Couldn't figure out how to get more accuracy from the trendline-generated equation.

Thanks for the help!

[shg]

Please Login or Register  to view this content.

Select A1:C1, paste the formula in the formula bar without the outermost curly braces, press and hold the Ctrl and Shift keys, then press Enter.