Plot least squares method

Question

So this is my octave plot of a regression line using the least squares method.

is there a way to add like little arrows or lines that connect the dots with the actual line? Here is an example of how i want it to look like. Thanks

Answer 1

I don't think there's a specific function for plotting residuals, but doing it manually is fairly straightforward:

% Let's assume this is our model, predicting y from x
  model = @sin;

% Define the x domain, which will be used for plotting
  xdomain = 0:0.1:10;

% Define some (x,y) input points
  xpoints = 10 * rand(1, 10);
  ypoints = model(xpoints) + 0.5 * randn(size(xpoints));

% Plot model
  plot( xdomain, model(xdomain), 'k-', 'linewidth', 1.5 );
  hold on;

% Plot input points
  plot( xpoints, ypoints, 'ko', 'markersize', 8, 'markeredgecolor', 'k', 'markerfacecolor', [0.4,0.4,0.4], 'linewidth', 1.5 )

% Plot residual lines
  plot( [xpoints;xpoints], [model(xpoints);ypoints], 'k:', 'linewidth', 1.5 )
  hold off;