root/MachineLearningWithMATLAB/Regression_FuelEconomy/showFit.m @ 11
| 10 | anderm8 | function showFit(yactual,yfit)
|
|
plotFit(yactual, yfit)
|
|||
R2_Final = R2(yactual,yfit);
|
|||
fprintf('\nTest Data R^2: %f\n',R2_Final);
|
|||
end
|
|||
function R_Sqr = R2(Y, YHat )
|
|||
%RSQUARED
|
|||
SStot = sum((Y - mean(Y)).*(Y - mean(Y)));
|
|||
% Calculate residuals
|
|||
resid = Y - YHat;
|
|||
% Square residuals
|
|||
resid_sqrd = resid.*resid;
|
|||
% Take the sum of the squared residuals
|
|||
SSerr = sum(resid_sqrd);
|
|||
% Calculate R^2
|
|||
R_Sqr = 1 - (SSerr/SStot);
|
|||
end
|
|||
function plotFit(y,varargin)
|
|||
% Plot fit
|
|||
% Copyright 2015 The MathWorks, Inc.
|
|||
figure
|
|||
legendName = cell(length(varargin),1);
|
|||
hold all
|
|||
for ii = 1:length(varargin)
|
|||
yfit = varargin{ii};
|
|||
scatter(yfit, (y- yfit)./yfit,'o','filled')
|
|||
rs = R2(y,yfit);
|
|||
legendName{ii,1} = sprintf('%s (R^2=%0.3f)', inputname(ii+1), rs);
|
|||
end
|
|||
ylim([-1.5 1.5])
|
|||
hold off
|
|||
legend(legendName)
|
|||
refline(0,0)
|
|||
xlabel('Estimated Fuel Economy (MPG)')
|
|||
ylabel('Error/Estimated Fuel Economy')
|
|||
figure;
|
|||
hold all
|
|||
axis square
|
|||
linesymbols = {'+','o','*','x'};
|
|||
for ii = 1:length(varargin)
|
|||
counter = 1 + mod((ii-1),4);
|
|||
yfit = varargin{ii};
|
|||
plot(y, yfit,linesymbols{counter})
|
|||
end
|
|||
hold off
|
|||
line([0 max(y)] ,[0 max(y)],'Color','k')
|
|||
xlabel('Actual Fuel Economy (MPG)')
|
|||
ylabel('Estimated Fuel Economy (MPG)')
|
|||
legend(legendName)
|
|||
end
|