My Math Forum Computing the error of a function of the fit parameters

 Probability and Statistics Basic Probability and Statistics Math Forum

 August 14th, 2018, 01:34 AM #1 Newbie   Joined: Aug 2018 From: Frankfurt Posts: 1 Thanks: 0 Computing the error of a function of the fit parameters Hi, I'm using the Finite Center of Rotation method described at the end of this thread to estimate the center of rotation and angle between two vectors A, B of corresponding points in the plane. Effectively, it is just a least-squares method with complex numbers to find the x which best minimizes |A - Bx| = (A-Bx)* (A-Bx) . The best-fit rotation is then simply arg(x)=arctan(imag(x)/real(x)). So far so good. Now I also want a measure of quality in terms of rotation, i.e. as an error of the angle, and I'm not sure how to achieve that from the pseudo-inverse of B, the vector of residuals or otherwise. Any idea ?

 Tags computing, error, fit, function, parameters

 Thread Tools Display Modes Linear Mode

 Similar Threads Thread Thread Starter Forum Replies Last Post marly892 Math 0 July 20th, 2017 01:17 AM JJeter2 Pre-Calculus 8 August 27th, 2015 11:12 PM fysmat Applied Math 3 April 1st, 2015 11:00 AM fysmat Applied Math 7 March 1st, 2014 08:17 AM OriaG Calculus 5 July 31st, 2012 02:57 AM

 Contact - Home - Forums - Cryptocurrency Forum - Top