plotroc.m

Matlab工具包补充算发

function [falseAlarmRate, detectionRate] = plotROC(confidence, testClass, col)
% You pass the scores and the classes, and the function returns the false
% alarm rate and the detection rate for different points across the ROC.
%
% [faR, dR] = plotROC(score, class)
%
% it generates 150 points. 
%  faR (false alarm rate) is uniformly sampled from 0 to 1
%  dR (detection rate) is computed using the scores.
%
% class = 0 => target absent
% class = 1 => target present
%
% score is the output of the detector, or any other measure of detection.
% There is not plot unless you add a third parameter that is the color of
% the graph. For instance:
% [faR, dR] = plotROC(score, class, 'r')

confidence = confidence(:);
testClass = testClass(:);
ndxAbs = find(testClass==0); % absent
ndxPres = find(testClass==1); % present

[th, j] = sort(confidence(ndxAbs));
th = th(fix(linspace(1, length(th), 150))); % here the number of points is hardcoded to be 150.

for t=1:length(th)
  detectionRate(t)  = sum(confidence(ndxPres)>=th(t)) / length(ndxPres);
  falseAlarmRate(t) = sum(confidence(ndxAbs)>=th(t)) / length(ndxAbs);
  %detections(t)     = sum(confidence(ndxPres)>=th(t));
  %falseAlarms(t)    = sum(confidence(ndxAbs)>=th(t));
end

if nargin == 3
    plot(falseAlarmRate, detectionRate, [col 'o-']); axis([0 1 0 1])
    %loglog(falseAlarmRate, detectionRate, [col '-']); 

    grid on
    ylabel('detection rate')
    xlabel('false alarm rate')
end