How to apply linear least squares fit for two entities?

Dear Researchers,

I have two entities namely sigma (dB) and U (m/s) respectively. As given in attach figure, a linear least squares is used to fit among two entities. I have attached .mat files which contains sigma and U values. I need to fit this curve to analyse Bias and RMS.

I lookforward to receive your valubale suggestions.

Thank you!

[dpb]

That certainly doesn't look to be nonlinear. What's the functional form you think you should be fitting (a literature reference to the type of data/from whencecame the above figure, maybe???)

[Amjad Iqbal]

As a matter of fact, I have sigma_VH from product and using equation-1 given in reference to obtain U.

In Fig-1 (d), this fit is applied among Sigma_VH and U. According to what they claim is nonlinear. However, I tired is linear. As you can see in attached result.

What's your suggetions and how to fit line among these two?

Answer 1

[Matt J]

You can do a total least squares fit using linear2dFit() from,

https://www.mathworks.com/matlabcentral/fileexchange/87584-object-oriented-tools-for-fitting-conics-and-quadrics

[Amjad Iqbal]

@Matt J

I tried but this doesn't apply, if you may chech with my attached datasets will be helpful.

[Matt J]

Works fine for me

>> fobj=linear2dFit([U_vh1(:)';sigma_o_VH1(:).'])

fobj =

linear2dFit with properties:

p1: [4.7831 -27.4053]

p2: [9.1807 -19.8232]