How can I fit data which is like that it will check directly several possible issue of fitting then in output it will give right fitted curve ?

I have X and Y data and want to plot and then want to fit like that it can give result which is having nearest possible fit. As suppose first of all it will check first order polynomial then second order, and in output it have to give the which is good enough fit?

Answer 1

[Shashank Prasanna]

If you are looking to do this automatically you can use Stepwise regression.

http://www.mathworks.com/help/stats/linearmodel.stepwise.html

specify the 'polyijk' model spec. The function will choose the best model based on a specified criterion.

[RS]

2.07948  0.967031
2.08032  0.966973
2.08115  0.96657
2.08199  0.966362
2.08283  0.966502
2.08366  0.96667
2.0845  0.96647
2.08534  0.966484
2.08618  0.966465
2.08702  0.966699
2.08786  0.966858
2.0887  0.966567
2.08954  0.966397
2.09038  0.966323
I have this data? How can I fit using polyijk?

[Shashank Prasanna]

Did you make an attempt first? There are examples in the link I gave you.