% Three peaks, noise test
% Demonstration script for peakfit.m, with built-in signal generator.  
% You can change any of the initial values in lines 12-28.
% T. C. O'Haver (toh@umd.edu).  Version 4: May, 2009. 
clear
close
format long g
format compact
warning off all

% Initial settings of users-adjustable parameters
n=300;  % maximum x-value
increment=1; % Difference between adjacent x values
x=[1:increment:n];  % Creates x-vector
amp=[1 2 3];  % Amplitudes of the peaks (one entry for each peak)
pos=[100 130 200];   % Positions of the peaks (one entry for each peak)
wid=[30 40 50];   % Widths of the peaks (one entry for each peak)
noise=.01; % Amplitude of random noise added to signal
NumRepeats=2; % Number of repeat samples with independent noise
NumTrials=1; % Number of trail fits per sample
PeakShape=1; % Gaussian=1, Lorentzian=2, Logistic=3, Pearson=4
% "extra" determines the shape of the Pearson function. When extra is
% 1.0, the shape is Lorentzian. When extra is large (>20), the shape 
% approaches Gaussian. At small values (<1), the shape is a pointy "cusp".
extra=2; 
center=160; % x-value in center of fitted range
window=260; % range of x-values in fitted range
startvector=[98 28 129 41 190 49]; % First guess vector [pos1 wid1 pos2 wid2 ...] Must be 2 for each peak

% Generate simulated signal peaks
% A = matrix containing one of the unit-amplidude functions in each of its rows
clear A
A=zeros(length(pos),length(x));
for k=1:length(pos),
  switch PeakShape
  case 1
    A(k,:)=exp(-((x-pos(k))./(0.6006.*wid(k))) .^2);  % Gaussian function
  case 2
    A(k,:)=ones(size(x))./(1+((x-pos(k))./(0.5.*wid(k))).^2); % Lorentzian function
  case 3
    L = exp(-((x-pos(k))/(.477.*wid(k))) .^2);
    A(k,:)= (2.*L)./(1+L);
  case 4
      % pearson function (shape determined by value of extra in line 21
      A(k,:)=ones(size(x))./(1+((x-pos(k))./((0.5.^(2/extra)).*4.62.*wid(k))).^2).^extra;
  end % switch
  area(k)=amp(k).*trapz(x,A(k,:)); % Computes peak areas via trapaziod method
end % for
TrueY=amp*A;  % Multiplies each row by the corresponding amplitude and adds them up
% To test the autozero baseline subtraction, un-comment next line.
% TrueY=TrueY+ones(size(x))./(1+((x+100)./(300)).^2); % Adds sloping curved baseline

% Arrange the peak parameters of the simulated signal into a matrix
ColumnLabels='           Peak#   Position     Height     Width       Area';
NumPeaks=length(pos);
for m=1:NumPeaks,
  if m==1,  
     ActualParameters=[m pos(m) amp(m) wid(m) area(m) ];, 
  else
     ActualParameters=[ActualParameters; [m pos(m) amp(m) wid(m) area(m) ]];
  end
end

% Generate and fit NumRepeats separate samples of the simulated signal with noise.
AllFitResults=zeros(NumPeaks,5,NumRepeats);
AllFittingErrors=zeros(1,NumRepeats);
AllPercentParameterErrors=zeros(NumPeaks,5,NumRepeats);
for repeat=1:NumRepeats,
   %  Add random noise to simulated signal
   y = TrueY + noise .* randn(size(x));
   % Execute the peakfit script
   [FitResults,LowestError]=peakfit([x' y'],center,window,NumPeaks,PeakShape,extra,NumTrials,startvector); 
   AllFitResults(:,:,repeat)=FitResults;
   AllFittingErrors(repeat)=LowestError;
   AllPercentParameterErrors(:,:,repeat)=100.*(FitResults-ActualParameters)./ActualParameters;
end

% Calculate the mean and relative standard deviaiton of parameters
for peak=1:NumPeaks,
      MeanParameters(peak,:)=mean(squeeze(AllFitResults(peak,:,:))');
      RSDparameters(peak,:)=100.*std(squeeze(AllFitResults(peak,:,:))')./mean(squeeze(AllFitResults(peak,:,:))');
end
  disp('-------------------------------------------------------')
  disp(['Noise amplitude: ' num2str(noise)])
  disp(['Number of repeats samples: ' num2str(NumRepeats)])
  disp(['Number of trial fits per sample: ' num2str(NumTrials)])
  disp('Peak#    Position        Height       Width       Area');
  disp('Actual parameters');disp(num2str(ActualParameters))
  disp('Average measured parameters');disp(num2str(MeanParameters))
  RSDparameters(:,1)=ActualParameters(:,1);
  disp('Percent RSD of measured parameters');disp(num2str(RSDparameters))
  meanRSD=mean(RSDparameters);
  meanRSD=mean(meanRSD(2:4));
   disp(['Average %RSD of measured parameters: ' num2str(meanRSD)])
  disp(['Average percent fitting error: ' num2str(mean(AllFittingErrors))])
  PercentParameterErrors=100.*(MeanParameters-ActualParameters)./ActualParameters;
  PercentParameterErrors(:,1)=ActualParameters(:,1);
  disp('Percent error of measured parameters');disp(num2str(PercentParameterErrors))
  AverageParameterErrors=mean(abs(PercentParameterErrors(:,2:5)))
  disp(['Overall Average Parameter Error: ' num2str(mean(AverageParameterErrors)) ])