Question

P2) Write a MATLAB function myLinReg Username that is defined below. Use the in-built MATLAB function sum( ) to ease your programming (you are NOT allowed to use polyfit() and polyval()). Test the output of this function with the results obtained by hand for P1. Submit printout of program and a screenshot of the command window showing the results for P1. function [coeffvec, r2]-myLinReg_Username (xi,yi) %Function file: myLinReg-Username .m Purpose : 8 To obtain the parameters of a L.S. linear best-fit for 8 given data eRecord of revisions (Date | Programmer | Change) t Date Name original program Main variables : % INPUTS : s xi 8 yi - (1xn) vector of given x values - (1xn) vector of given y values OUTPUTS % coeffvec- (2x1) vector of coefficients of linear best-fit a0+a1*x coeffvec (1) is the intercept (a0) coeffVec (2) is the slope (al)

Answer 1

Required Matlab code with explanatory comments is given below:

${\color{Red} \text{START CODE}}$

function [coeffVec,r2] = Q_myLinReg(xi,yi)
%FINDS THE BEST FIT LINE PARAMETERS AND R-SQUARED VALUE
slope=sum((xi-mean(xi)).*(yi-mean(yi)))/sum(( xi-mean(xi)).^2); %formula for slope
intercept=mean(yi)-slope*mean(xi); %formula for intercept
coeffVec = [intercept,slope];
r = sum((xi-mean(xi)).*(yi-mean(yi)))/sqrt(sum((xi-mean(xi)).^2)*sum((yi-mean(yi)).^2)); %r valu
r2=r^2; %r squared value
end

${\color{Red} \text{END CODE}}$

Sample usage:

> [coeff,rsql-o myLinReg ([1,2,3], 04,5,5]) coeff = 3.6667 0. 0.5000 AKI 0.7500