fitplot

% With no input arguments, FITPLOT will plot a linear fit to each "line"
% data series that has a "marker" in the current axes. It will also put
% the equation describing the fit and the correlation coefficient in the
% upper left corner of the plot. NaNs are removed from the data. A flag
% is added to the equation if NaNs were present.
%
% FITPLOT(H) where H is a handle to a data series, will plot a linear fit
% to the data series. H can be a vector with multiple handles.
%
% FITPLOT(...,'robust') uses a robust fit instead of a regular fit.
%
% FITPLOT could be extended to handle other types of curves (polynomials,
% exponentials, etc).
%
% Example:
% x=(1:100)';
% a=-3*x+400+100*rand(100,1); %Noisy data with NaNs
% a([10 32 56])=nan;
% b=2*x+20; %Simple data with a perfect correlation
% c=10*x-200+50*randn(100,1); %More noisy data
% figure
% plot(x,a,'.',x,b,'r.',x,c,'g.')
% fitplot