Fit a line to data using regress

Hey all, I am trying to use a new matlab function "regress" and would like some help on how to use it to fit a line.

below is my code that I have so far, but lost on how to use the "regress" function.

x = 0:0.1:10;
n = 0;
noise = n*rand(1,length(x));
y = 2*x+1+noise;

figure(1);
plot(x,y,'d') 

Answer 1

[Adam Danz]

b = regress(y,x)

If x and y are vectors, they must be column vectors

regress(y(:),x(:))

For alternatives, see

https://www.mathworks.com/matlabcentral/answers/277495-how-to-fit-linear-regression-models-to-a-scatter-graph#answer_492643

[Nicholas Deosaran]

Oh okay, I see.

and to plot that line on the axis..will it be

x = 0:0.1:10;
n = 0;
noise = n*rand(1,length(x));
y = 2*x+1+noise;
b = regress(y(:),x(:));

plot(x,y,'d') %plot as diamonds
hold on 
plot(b,y)
hold off

is that correct?

[Adam Danz]

No, b is just the slope if you've used the syntax from my answer.

If you want to plot the regression line you'll need the y-intercept term. To get the y-intercept term you need to add a column of 1s to x.

b = regress(y(:),[ones(size(x(:))),x(:)]); 

b is a 2x1 vector containing the [intercept; slope];

Use refline(b(2),b(1)) to add the regression line.

Or, check out lsline.

Or, check out polyfit(x,y,1) along with refline.