Introduce Deming/Orthogonal regression

[JesperGr]

Deming/Orthogonal regression, least-Squares fitting the points in the 2D dataset (x,y) to a line a*x + b*y + c = 0 For delta equal 1 (the default value), this is performing orthogonal regression, minimizing the sum of squared perpendicular distances from the data points to the regression line.

Discussed/suggested here: https://github.com/mathnet/mathnet-numerics/discussions/786

Questions:

• Is Fit.Line2D a good name?
• Where to put StandardLineToYxLine?

[jkalias]

• Is Fit.Line2D a good name?

I think that Fit.Deming would be more appropriate. Line2D seems like it performs the "usual" least squares fitting

• Where to put StandardLineToYxLine?

This is used purely for testing purposes, I think it is more appropriate in the testing class FitTests.StandardLineToYxLine

[JesperGr]

I like the Line in the name, because it is very similar to the SimpleRegression. But I am fine with Deming as well, so I have updated the Fit.Line2D to Fit.Deming. And moved the StandardLineToYxLine to the FitTests - I will then just keep my own line tools alive.