How to Calculate Uncertainty in Polynomial Fit Parameters Using Excel?

[JamesJames]

I performed a fit, using EXCEL, to data and the fit gave me a polynomial. How do I get the uncertainty in the resulting parameters. For example the least squares fit gives y = Ax+B and there are equations to give the uncertainties in A and B. IS there a program that does this for non linear fits?

James

 

[dduardo]

You'll want to output the R^2 correlation coefficient. The closer R^2 is to 1, the more accurate the equation is. I know excel can output it, but I'm not sure exactly how since I haven't used excel in a long time.

 

[FredGarvin]

In the "INSERT TRENDLINE" dialog box, there is a tab for options. At the bottom of the dialog box you will see three check boxes. "SET INTERCEPT TO 0," "DISPLAY EQUATION ON CHART" and "DISPLAY R^2 ON CHART." Select the ones you want.

Word of warning, if you check the equation output, you may have to increase the number of decimal places shown in the equation to get better correlation between your regression and the data.

 

[JamesJames]

How about performing a fit to data that should obey [csc(theta/2)]^4. Would this be out of a numerical methods book or something?

James