I've take a large amount of data (200+ runs) and used the Compare Runs tool to get Average, StdDev, 3Sigma stats out of it. I get a pretty decent looking fit when I use the Curve Fit with 2nd order to simplify the Average, StdDev and 3Sigma. Where can I get the equations and R-squared for the fitted curve? I can export them to Excel and get it but that is pretty tedious...
Solved! Go to Solution.
In the front/back display you can right click to add double X cursor as shown below:
once the cursor was added then right click on the cursor as shown below:
make sure Snap to data values is selected then go to Calculations to select Linear Regression and the a, b & R^2 values are shown in the curve legend. Also curve legend values can be copy to excel as well.
If you want calculation done for the whole frequency range, please move left cursor to zero hertz and right cursor to the maximum bandwidth.
I missed read the questions, my screen shots and answer are for first order curve fitted not second order.
Thanks for the instructions on how to get this info!
I'm a little confused on the results though. I place the curves I want the regression for and then click the Curve Fitting button in the Navigator toolbar and select 2. I placed the same curves back in on top to visually compare the fit. The table with the linear regression is showing me what I think are resonable values for the original data and the fitted curves have a different equations. The a values are nearly exactly the same, b are pretty close.
The main question I have is the R-square. Is it reasonable to expect that a linear regression of a linear regression would have an R-square of basically 1? In two cases, it's less than 0.9. I'm not sure if this matters. The results are different than I was expecting so looking for some education here.
I'm very sorry, in the previous replied my instruction was for first order curve fitted not second order curve fitted. We are looking into how to provide R^2 for 2nd curve fitted in Testlab.