Test FAMS on 1D data

Contents

randomize data

max_uniform=100;
vect1 = rand(1000,1)*max_uniform; % uniform
vect2 = randn(100,1)*3+20; %first gauss
vect3 = randn(100,1)*3+60; %second gauss
data=[vect1;  vect2; vect3];

run FAMS

[modes,p_modes,w_modes ] =fams(data,20,15,100,'exmpl1D','res/','-h',10);

Load data points from matlab ...done
RunFAMS with res/pilot_100_exmpl1D.txt ...
Running FAMS with K=20 L=15
 Run pilot fixed bandwith...done.
 Start MS iterations..........done.
 Join Modes with adaptive h/2, min pt=40, jump=1
            pass 1.done
            pass 2..........nrel 6
0 0
1 1
2 2
3 3
4 4
5 5
done
Save convergence points   ...done
Save joined convergence points ...done
Save indicies of modes  ...done
FAMS done.

view results

p_modes
figure(1),clf;
subplot(2,2,1),plot(data,'*'),title('1D data (x:sample num, y:value)');
subplot(2,2,2),plot(modes,'*'),title('after fams');
[f,xi]=ksdensity(modes,'width',2);
subplot(2,2,3),plot(xi,f,'.g');
subplot(2,2,4),hist(modes,10);

% p_modes =
%   420.0000   19.9799
%   386.0000   59.2612
%   121.0000   43.4002
%   114.0000   83.6234
%    88.0000   92.7286
%    71.0000   75.0310

p_modes =
  387.0000   59.6318
  367.0000   20.2818
  133.0000   92.9923
  124.0000   45.1230
   99.0000   83.5047
   90.0000    4.6766

Published with MATLAB® 7.0.1