Mentions légales du service

Skip to content
Snippets Groups Projects
Commit 890a62d5 authored by hhakim's avatar hhakim
Browse files

Integrate bsl matlab demo scripts into matfaust.demo namespace. 

Also updating API doc relative to these scripts and removing useless variables and commands in cmake scripts (since we moved some demos).
parent e5d58eec
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
Showing
with 417 additions and 427 deletions
CMakeLists.txt
+ 2
− 21
View file @ 890a62d5
......@@ -372,27 +372,16 @@ if (BUILD_WRAPPER_MATLAB)
########## DEMO DIRECTORIES #############
###### SRC DIRECTORY #####
set(FAUST_DEMO_SRC_DIR ${FAUST_MISC_DIR}/demo CACHE INTERNAL "")
set(FAUST_DEMO_BSL_SRC_DIR ${FAUST_DEMO_SRC_DIR}/Brain_source_localization CACHE INTERNAL "")
#set(FAUST_DEMO_BSL_DATA_SRC_DIR ${FAUST_DEMO_SRC_DIR}/Brain_source_localization/data/ CACHE INTERNAL "")
set(FAUST_DEMO_TOOLS_SRC_DIR ${FAUST_DEMO_SRC_DIR}/tools CACHE INTERNAL "")
set(FAUST_DEMO_HADAMARD_SRC_DIR ${FAUST_DEMO_SRC_DIR}/Hadamard_factorization CACHE INTERNAL "")
set(FAUST_DEMO_FFT_SRC_DIR ${FAUST_DEMO_SRC_DIR}/FFT CACHE INTERNAL "")
set(FAUST_DEMO_TIMECOMP_SRC_DIR ${FAUST_DEMO_SRC_DIR}/Runtime_comparison CACHE INTERNAL "")
set(FAUST_DEMO_QUICKSTART_SRC_DIR ${FAUST_DEMO_SRC_DIR}/Quick_start CACHE INTERNAL "")
###### BIN DIRECTORY ######
set(FAUST_DEMO_BIN_DIR ${FAUST_MATLAB_BIN_DIR}/demo CACHE INTERNAL "")
set(FAUST_DEMO_BSL_BIN_DIR ${FAUST_DEMO_BIN_DIR}/Brain_source_localization CACHE INTERNAL "")
set(FAUST_DEMO_TOOLS_BIN_DIR ${FAUST_DEMO_BIN_DIR}/tools CACHE INTERNAL "")
set(FAUST_DEMO_HADAMARD_BIN_DIR ${FAUST_DEMO_BIN_DIR}/Hadamard_factorization CACHE INTERNAL "")
set(FAUST_DEMO_FFT_BIN_DIR ${FAUST_DEMO_BIN_DIR}/FFT CACHE INTERNAL "")
set(FAUST_DEMO_BSL_DATA_BIN_DIR ${FAUST_DEMO_BSL_BIN_DIR}/data/ CACHE INTERNAL "")
set(FAUST_DEMO_BSL_OUTPUT_BIN_DIR ${FAUST_DEMO_BSL_BIN_DIR}/output/ CACHE INTERNAL "")
set(FAUST_DEMO_HADAMARD_OUTPUT_BIN_DIR ${FAUST_DEMO_HADAMARD_BIN_DIR}/output/ CACHE INTERNAL "")
set(FAUST_DEMO_TIMECOMP_BIN_DIR ${FAUST_DEMO_BIN_DIR}/Runtime_comparison CACHE INTERNAL "")
set(FAUST_DEMO_FIG_BIN_DIR ${FAUST_DEMO_BIN_DIR}/Figures CACHE INTERNAL "")
set(FAUST_DEMO_QUICKSTART_BIN_DIR ${FAUST_DEMO_BIN_DIR}/Quick_start CACHE INTERNAL "")
set(FAUST_MATFAUST_DEMO_DATA_BIN_DIR ${FAUST_MATLAB_BIN_DIR}/data CACHE INTERNAL "")
#### INSTALL DIRECTORY ####
# No need to specify installation demo directory because
......@@ -457,14 +446,6 @@ set(INSTALL_DIR_PERMISSION OWNER_READ OWNER_WRITE OWNER_EXECUTE GROUP_READ GROUP
set(INSTALL_FILE_PERMISSION OWNER_READ OWNER_WRITE OWNER_EXECUTE GROUP_READ GROUP_WRITE GROUP_EXECUTE WORLD_READ WORLD_WRITE WORLD_EXECUTE)
if (BUILD_WRAPPER_MATLAB)
file(MAKE_DIRECTORY ${FAUST_DEMO_BSL_OUTPUT_BIN_DIR})
file(MAKE_DIRECTORY ${FAUST_DEMO_HADAMARD_OUTPUT_BIN_DIR})
file(MAKE_DIRECTORY ${FAUST_DEMO_FIG_BIN_DIR})
endif (BUILD_WRAPPER_MATLAB)
###### TARGETS ######
set(FAUST_LINEAR_OPERATOR_TARGET faust_linear_operator CACHE INTERNAL "")
set(FAUST_ALGORITHM_TARGET faust_algorithm CACHE INTERNAL "")
......
gen_doc/CMakeLists.txt
+ 1
− 1
View file @ 890a62d5
......@@ -7,7 +7,7 @@ if(BUILD_DOCUMENTATION)
string(CONCAT DOXYGEN_FILE_PATTERNS "*.cpp *.hpp *.h *.cu *.hu")
endif()
if(BUILD_WRAPPER_MATLAB)
string(CONCAT DOXYGEN_FILE_PATTERNS ${DOXYGEN_FILE_PATTERNS} " Faust.m StoppingCriterion.m ConstraintGeneric.m ConstraintMat.m ConstraintReal.m ConstraintInt.m ConstraintName.m ParamsFact.m ParamsHierarchicalFact.m ParamsPalm4MSA.m FaustFactory.m norm_hadamard.m demo_fact_hadamard.m speed_up_hadamard.m construct_Faust_from_factors.m quick_start.m factorize_matrix.m speed_up_fourier.m")
string(CONCAT DOXYGEN_FILE_PATTERNS ${DOXYGEN_FILE_PATTERNS} " Faust.m StoppingCriterion.m ConstraintGeneric.m ConstraintMat.m ConstraintReal.m ConstraintInt.m ConstraintName.m ParamsFact.m ParamsHierarchicalFact.m ParamsPalm4MSA.m FaustFactory.m norm_hadamard.m demo_fact_hadamard.m speed_up_hadamard.m construct_Faust_from_factors.m quick_start.m factorize_matrix.m speed_up_fourier.m Fig_BSL.m BSL.m")
endif()
if(BUILD_WRAPPER_PYTHON)
string(CONCAT DOXYGEN_FILE_PATTERNS ${DOXYGEN_FILE_PATTERNS} " pyfaust.py")
......
misc/demo/Brain_source_localization/BSL.m deleted 100755 → 0
+ 0
− 164
View file @ e5d58eec
%% Description BSL
% Brain source localization
%
% This script performs brain source localization using several gain
% matrices [2], including FAuSTs, and OMP solver. It reproduces the
% source localization experiment of [1].
% The results are stored in "./output/results_BSL_user.mat".
% DURATION: Computations should take around 3 minutes.
%
% The MEG gain matrices used are
% the precomputed ones in "./data/faust_MEG_rcg_X.mat"
%
% For more information on the FAuST Project, please visit the website of
% the project : <http://faust.inria.fr>
%
%% License:
% Copyright (2016): Nicolas Bellot, Adrien Leman, Thomas Gautrais, Luc Le Magoarou, Remi Gribonval
% INRIA Rennes, FRANCE
% http://www.inria.fr/
%
% The FAuST Toolbox is distributed under the terms of the GNU Affero
% General Public License.
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU Affero General Public License as published
% by the Free Software Foundation.
%
% This program is distributed in the hope that it will be useful, but
% WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
% See the GNU Affero General Public License for more details.
%
% You should have received a copy of the GNU Affero General Public License
% along with this program. If not, see <http://www.gnu.org/licenses/>.
%
%% Contacts:
% Nicolas Bellot : nicolas.bellot@inria.fr
% Adrien Leman : adrien.leman@inria.fr
% Thomas Gautrais : thomas.gautrais@inria.fr
% Luc Le Magoarou : luc.le-magoarou@inria.fr
% Remi Gribonval : remi.gribonval@inria.fr
%
%% References:
% [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
% approximations of matrices and applications", Journal of Selected
% Topics in Signal Processing, 2016.
% <https://hal.archives-ouvertes.fr/hal-01167948v1>
%
% [2] A. Gramfort, M. Luessi, E. Larson, D. Engemann, D. Strohmeier,
% C. Brodbeck, L. Parkkonen, M. Hamalainen, MNE software for processing
% MEG and EEG data <http://www.ncbi.nlm.nih.gov/pubmed/24161808>,
% NeuroImage, Volume 86, 1 February 2014, Pages 446-460, ISSN 1053-8119,
% [DOI] <http://dx.doi.org/10.1016/j.neuroimage.2013.10.027>
%%
import matfaust.Faust
runPath=which(mfilename);
pathname = fileparts(runPath);
% folder_path where the MEG and Faust approximations are stored
BSL_data_pathName=[pathname filesep 'data'];
% files where a precomputed FauST approximation of MEG matrix are stored
MEG_FAuST_list_filename={'faust_MEG_rcg_6.mat','faust_MEG_rcg_8.mat','faust_MEG_rcg_16.mat','faust_MEG_rcg_25.mat'};
nb_FAuST_MEG = length(MEG_FAuST_list_filename);
% list of the MEG matrices (the original one and her Faust approximations)
MEG_list = cell(1,nb_FAuST_MEG+1);
nb_MEG_matrix = length(MEG_list);
%% Loading the MEG matrix and the points in the brain
load([BSL_data_pathName filesep 'matrix_MEG.mat' ]);
MEG_matrix = normalizeCol(matrix');% normalization of the columns of the MEG matrix
MEG_list{1}=MEG_matrix;
%% Loading the precomputed FAuST representing the MEG matrix
% RCG (Relative Complexity Gain) : the theoretical speed-up using Faust
RCG_list=zeros(1,nb_FAuST_MEG);
for i=1:nb_FAuST_MEG
load([BSL_data_pathName filesep MEG_FAuST_list_filename{i}]);
facts = normalizeCol(facts,lambda); % normalization of the columns of the FAUST
MEG_FAuST=Faust(facts); % construct the FAuST from its factorscons15_row
MEG_list{i+1}=MEG_FAuST; % store the different FAuST approximationscons21_col
RCG_list(i)=rcg(MEG_FAuST); % compute the RCG of the given FAuST
end
M=size(MEG_matrix,2); % Number of points used in the MEG matrix
Ntraining = 500; % Number of training vectorscons22_row
Sparsity = 2; % Number of sources per training vector
dist_paliers = [0.01,0.05,0.08,0.5];
resDist = zeros(nb_MEG_matrix,numel(dist_paliers)-1,Sparsity,Ntraining);% (Matrice,dist_sources,src_nb,run);
compute_Times = zeros(nb_MEG_matrix,numel(dist_paliers)-1,Ntraining);
h = waitbar(0,['Brain Source Localization : MEG matrix and its faust approximations with omp solver']);
nb_palier=numel(dist_paliers)-1;
for k=1:nb_palier;
%Parameters settings
% generates the different source position
Gamma = zeros(M,Ntraining);
for ii=1:Ntraining
dist_sources = -1;
while ~((dist_paliers(k)<dist_sources)&&(dist_sources<dist_paliers(k+1)))
Gamma(:,ii) = sparse_coeffs(MEG_matrix, 1, Sparsity); %
idx = find(Gamma(:,ii));
dist_sources = norm(points(idx(1),:) - points(idx(2),:));
end
end
% compute the data registered by MEG sensor
Data = MEG_matrix*Gamma;
for i=1:Ntraining
waitbar(((k-1)*Ntraining+i)/(nb_palier*Ntraining));
% index of the real source localization
idx = find(Gamma(:,i));
for j=1:nb_MEG_matrix
MEG_FAuST = MEG_list{j};
% find the active source
tic
sol_solver=greed_omp_chol(Data(:,i),MEG_FAuST,M,'stopTol',1*Sparsity,'verbose',false);
compute_Times(j,k,i)=toc;
% compute the distance between the estimated source and the real one
idx_solver = find(sol_solver);
resDist(j,k,1,i) = min(norm(points(idx(1),:) - points(idx_solver(1),:)),norm(points(idx(1),:) - points(idx_solver(2),:)));
resDist(j,k,2,i) = min(norm(points(idx(2),:) - points(idx_solver(1),:)),norm(points(idx(2),:) - points(idx_solver(2),:)));
end
end
end
close(h);
matfile = fullfile(pathname, 'output/results_BSL_user');
save(matfile,'resDist','Sparsity','RCG_list','compute_Times','Ntraining','nb_MEG_matrix');
misc/demo/Brain_source_localization/Fig_BSL.m deleted 100755 → 0
+ 0
− 200
View file @ e5d58eec
%% Description Fig_BSL
% Brain source localization figures
% This script builds the BSL figure (Fig. 9.) used in [1].
%
% For more information on the FAuST Project, please visit the website of
% the project : <http://faust.inria.fr>
%
%% License:
% Copyright (2016): Nicolas Bellot, Adrien Leman, Thomas Gautrais, Luc Le Magoarou, Remi Gribonval
% INRIA Rennes, FRANCE
% http://www.inria.fr/
%
% The FAuST Toolbox is distributed under the terms of the GNU Affero
% General Public License.
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU Affero General Public License as published
% by the Free Software Foundation.
%
% This program is distributed in the hope that it will be useful, but
% WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
% See the GNU Affero General Public License for more details.
%
% You should have received a copy of the GNU Affero General Public License
% along with this program. If not, see <http://www.gnu.org/licenses/>.
%
%% Contacts:
% Nicolas Bellot : nicolas.bellot@inria.fr
% Adrien Leman : adrien.leman@inria.fr
% Thomas Gautrais : thomas.gautrais@inria.fr
% Luc Le Magoarou : luc.le-magoarou@inria.fr
% Remi Gribonval : remi.gribonval@inria.fr
%
%% References:
% [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
% approximations of matrices and applications", Journal of Selected
% Topics in Signal Processing, 2016.
% <https://hal.archives-ouvertes.fr/hal-01167948v1>
%
% [2] A. Gramfort, M. Luessi, E. Larson, D. Engemann, D. Strohmeier,
% C. Brodbeck, L. Parkkonen, M. Hamalainen, MNE software for processing
% MEG and EEG data <http://www.ncbi.nlm.nih.gov/pubmed/24161808>,
% NeuroImage, Volume 86, 1 February 2014, Pages 446-460, ISSN 1053-8119,
% [DOI] <http://dx.doi.org/10.1016/j.neuroimage.2013.10.027>
%%
runPath=which(mfilename);
pathname = fileparts(runPath);
matfile = fullfile(pathname, ['output' filesep 'results_BSL_user.mat']);
if (not(exist(matfile)))
error('run BSL.m before Fig_BSL.m');
end
%% figure configuration
figure_dir = [pathname filesep '..' filesep 'Figures'];
format_fig='-dpng';
load(matfile);
nb_approx_MEG=nb_MEG_matrix-1;
solver_choice='omp';
Ntest=Ntraining*Sparsity;
%% convergence analysis
d1 = cat(4,resDist(:,1,1,:),resDist(:,1,2,:));
d2 = cat(4,resDist(:,2,1,:),resDist(:,2,2,:));
d3 = cat(4,resDist(:,3,1,:),resDist(:,3,2,:));
test2 = 100*[squeeze(d1);zeros(1,Ntest);squeeze(d2);zeros(1,Ntest);squeeze(d3)];
f=figure('color',[1 1 1]);
T = bplot(test2','linewidth',1.5);
legend(T);
ylabel('Distance between true and estimated sources (cm)')
box on
ax = gca;
x_tick=[1:nb_approx_MEG+1;nb_approx_MEG+3:2*nb_approx_MEG+3;2*nb_approx_MEG+5:3*nb_approx_MEG+5];
mean_x_tick = round(mean(x_tick,2));
set(ax,'XTick',reshape(x_tick',1,numel(x_tick)));
set(ax,'xticklabel', [])
axi = axis;
axis([0 3*(nb_approx_MEG+2) -0.3 4.7])
HorizontalOffset = 10;
verticalOffset = 0.43;
verticalOffset2 = 0.7;
yTicks = get(ax,'ytick');
xTicks = get(ax, 'xtick');
minY = min(yTicks);
text(xTicks(1), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(2+nb_approx_MEG), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(1+2*(nb_approx_MEG+1)), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(1:(nb_approx_MEG+1)))), minY - verticalOffset2, '$1<d<5$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(nb_approx_MEG+2:2*(nb_approx_MEG+1)))), minY - verticalOffset2, '$5<d<8$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(2*(nb_approx_MEG+1)+1:3*(nb_approx_MEG+1)))), minY - verticalOffset2, '$d>8$','HorizontalAlignment','center','interpreter', 'latex');
for i=1:nb_approx_MEG
text(xTicks(i+1), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(i+2+nb_approx_MEG), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) ' }$' ],'HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(i+1+2*(nb_approx_MEG+1)), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) ' }$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
title(['BSL - convergence (C++ wrapper faust) ' solver_choice ' solver']);
f.Name =['Brain Source Localization : convergence with ' solver_choice '_solver (C++ wrapper)'];
figure_name=[figure_dir filesep 'BSL-convergence_Cpp_' solver_choice '_solver'];
print(figure_name, format_fig);
%% time comparison
timeS = cat(3,compute_Times(:,1,:),compute_Times(:,2,:),compute_Times(:,3,:));
timeS = squeeze(1000*timeS)';
f=figure('color',[1 1 1]);
T = bplot(timeS,'linewidth',1.5);
legend(T);
ylabel('Computed Time (ms)')
box on
ax = gca;
x_tick=[1:2*nb_approx_MEG+1];
mean_x_tick = round(mean(x_tick,2));
set(ax,'XTick',x_tick');
set(ax,'xticklabel', [])
axi = axis;
HorizontalOffset = 10;
verticalOffset = 0.5;
verticalOffset2 = 1;
yTicks = get(ax,'ytick');
xTicks = get(ax, 'xtick');
minY = min(yTicks);
text(xTicks(1), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
for i=1:nb_approx_MEG
text(xTicks(i+1), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
title(['BSL - time comparison (FAUST vs dense matrix) ' solver_choice ' solver']);
f.Name =['Brain Source Localization : time comparison with ' solver_choice 'solver'];
figure_name=[figure_dir filesep 'BSL-time_comparison_' solver_choice '_solver'];
print(figure_name, format_fig);
%% speed-up
mean_Times=mean(timeS);
dense_matrix_time=mean_Times(1);
real_RCG=dense_matrix_time./mean_Times;
f=figure;
ax = gca;
set(ax,'xticklabel', [])
set(ax,'Visible','off');
plot(1:nb_approx_MEG,real_RCG(2:end),'linewidth',1.5);
hold on
plot(1:nb_approx_MEG,ones(1,nb_approx_MEG),'linewidth',1.5);
verticalOffset=1.5;
minY=min([0.9,real_RCG(2:end)]);
maxY=max([0.9,real_RCG(2:end)]);
axis([1,nb_approx_MEG,minY,maxY]);
set(ax,'XTick',[]);
for i=1:nb_approx_MEG
text(i, minY -(maxY-minY)/20, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
legend('speed up FAuST','neutral speed up');
title(['BSL - speed up using FAUST ' solver_choice ' solver']);
f.Name =['Brain Source Localization : speed-up Faust with ' solver_choice 'solver'];
figure_name = [figure_dir filesep 'BSL-speed_up_' solver_choice ' solver'];
print(figure_name, format_fig);
%% console display
disp(['**** MEG with OMP solver time comparison ****']);
disp(['M tps : ' num2str(mean(timeS(:,1))) ]);
for i=1:nb_approx_MEG
disp([ 'M_' int2str(RCG_list(i)) ' tps : ' num2str(mean(timeS(:,i+1))) ', speed-up : ' num2str(real_RCG(i+1))]);
end
misc/demo/CMakeLists.txt
+ 11
− 41
View file @ 890a62d5
......@@ -2,70 +2,40 @@
file(GLOB MATLAB_SCRIPTS "*.m")
if(BUILD_WRAPPER_MATLAB)
configure_file(${FAUST_DEMO_SRC_DIR}/run_all_demo.m ${FAUST_DEMO_BIN_DIR}/run_all_demo.m COPYONLY)
############# Brain source localization demo ####################
file(GLOB BSL_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_BSL_SRC_DIR} "${FAUST_DEMO_BSL_SRC_DIR}/*.m")
foreach(matlab_script ${BSL_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_BSL_SRC_DIR}/${matlab_script} ${FAUST_DEMO_BSL_BIN_DIR}/${matlab_script} COPYONLY)
endforeach()
# copy BSL data matrices
file(GLOB BSL_DATA_FILE RELATIVE ${FAUST_DATA_MAT_DIR} "${FAUST_DATA_MAT_DIR}/*MEG*.mat")
foreach(data_BSL ${BSL_DATA_FILE})
configure_file(${FAUST_DATA_MAT_DIR}/${data_BSL} ${FAUST_DEMO_BSL_DATA_BIN_DIR}/${data_BSL} COPYONLY)
configure_file(${FAUST_DATA_MAT_DIR}/${data_BSL} ${FAUST_MATFAUST_DEMO_DATA_BIN_DIR}/${data_BSL} COPYONLY)
endforeach()
######### Hadamard factorization ###############################
file(GLOB HADAMARD_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_HADAMARD_SRC_DIR} "${FAUST_DEMO_HADAMARD_SRC_DIR}/*.m")
foreach(matlab_script ${HADAMARD_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_HADAMARD_SRC_DIR}/${matlab_script} ${FAUST_DEMO_HADAMARD_BIN_DIR}/${matlab_script} COPYONLY)
endforeach()
######### FFT (Fourier) speed-up ###############################
file(GLOB FFT_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_FFT_SRC_DIR} "${FAUST_DEMO_FFT_SRC_DIR}/*.m")
foreach(matlab_script ${FFT_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_FFT_SRC_DIR}/${matlab_script} ${FAUST_DEMO_FFT_BIN_DIR}/${matlab_script} COPYONLY)
endforeach()
####### Runtime Comparison ######################################
file(GLOB TIMECOMP_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_TIMECOMP_SRC_DIR} "${FAUST_DEMO_TIMECOMP_SRC_DIR}/*.m")
foreach(matlab_script ${TIMECOMP_MATLAB_SCRIPTS})
foreach(matlab_script ${TIMECOMP_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_TIMECOMP_SRC_DIR}/${matlab_script} ${FAUST_DEMO_TIMECOMP_BIN_DIR}/${matlab_script} COPYONLY)
endforeach()
##### Quick Start ##############################################
file(GLOB TOOLS_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_QUICKSTART_SRC_DIR} "${FAUST_DEMO_QUICKSTART_SRC_DIR}/*.m")
foreach(tools ${TOOLS_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_QUICKSTART_SRC_DIR}/${tools} ${FAUST_DEMO_QUICKSTART_BIN_DIR}/${tools} COPYONLY)
# copy data matrix for quick_start.m
configure_file(${FAUST_DATA_MAT_DIR}/faust_quick_start.mat ${FAUST_MATFAUST_DEMO_DATA_BIN_DIR}/faust_quick_start.mat COPYONLY)
endforeach()
configure_file(${FAUST_DATA_MAT_DIR}/faust_quick_start.mat ${FAUST_DEMO_QUICKSTART_BIN_DIR}/faust_quick_start.mat COPYONLY)
#################### Tools directory ############################
file(GLOB TOOLS_MATLAB_SCRIPTS RELATIVE ${FAUST_DEMO_TOOLS_SRC_DIR} "${FAUST_DEMO_TOOLS_SRC_DIR}/*.m")
foreach(tools ${TOOLS_MATLAB_SCRIPTS})
configure_file(${FAUST_DEMO_TOOLS_SRC_DIR}/${tools} ${FAUST_DEMO_TOOLS_BIN_DIR}/${tools} COPYONLY)
endforeach()
#the installation is made in wrapper/matlab/CmakeList.txt as the demo is now Matlab code
#the installation is made in wrapper/matlab/CmakeList.txt as the demo is now Matlab code
endif()
misc/demo/run_all_demo.m
+ 2
− 0
View file @ 890a62d5
......@@ -59,6 +59,7 @@ construct_Faust_from_factors;
%% brain source localization
disp('*********** Brain Source Localization *************');
import matfaust.demo.bsl.*
BSL;
Fig_BSL;
......@@ -72,6 +73,7 @@ norm_hadamard;
%% Fourier speed-up
disp('*********** Fourier speed-up *************');
iport matfaust.demo.fft.*
speed_up_fourier;
......
wrapper/matlab/+matfaust/+demo/+bsl/BSL.m 0 → 100755
+ 186
− 0
View file @ 890a62d5
%% Description BSL
% Brain source localization
%
% This script performs brain source localization using several gain
% matrices [2], including FAuSTs, and OMP solver. It reproduces the
% source localization experiment of [1].
% The results are stored in "./output/results_BSL_user.mat".
% DURATION: Computations should take around 3 minutes.
%
% The MEG gain matrices used are
% the precomputed ones in "./data/faust_MEG_rcg_X.mat"
%
% For more information on the FAuST Project, please visit the website of
% the project : <http://faust.inria.fr>
%
%% License:
% Copyright (2016): Nicolas Bellot, Adrien Leman, Thomas Gautrais, Luc Le Magoarou, Remi Gribonval
% INRIA Rennes, FRANCE
% http://www.inria.fr/
%
% The FAuST Toolbox is distributed under the terms of the GNU Affero
% General Public License.
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU Affero General Public License as published
% by the Free Software Foundation.
%
% This program is distributed in the hope that it will be useful, but
% WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
% See the GNU Affero General Public License for more details.
%
% You should have received a copy of the GNU Affero General Public License
% along with this program. If not, see <http://www.gnu.org/licenses/>.
%
%% Contacts:
% Nicolas Bellot : nicolas.bellot@inria.fr
% Adrien Leman : adrien.leman@inria.fr
% Thomas Gautrais : thomas.gautrais@inria.fr
% Luc Le Magoarou : luc.le-magoarou@inria.fr
% Remi Gribonval : remi.gribonval@inria.fr
%
%% References:
% [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
% approximations of matrices and applications", Journal of Selected
% Topics in Signal Processing, 2016.
% <https://hal.archives-ouvertes.fr/hal-01167948v1>
%
% [2] A. Gramfort, M. Luessi, E. Larson, D. Engemann, D. Strohmeier,
% C. Brodbeck, L. Parkkonen, M. Hamalainen, MNE software for processing
% MEG and EEG data <http://www.ncbi.nlm.nih.gov/pubmed/24161808>,
% NeuroImage, Volume 86, 1 February 2014, Pages 446-460, ISSN 1053-8119,
% [DOI] <http://dx.doi.org/10.1016/j.neuroimage.2013.10.027>
%%
%===============================================================================
%> Brain source localization
%===
%>
%> This script performs brain source localization using several gain
%> matrices <a href="http://www.ncbi.nlm.nih.gov/pubmed/24161808">[2]</a>, including FAuSTs, and OMP solver. It reproduces the
%> source localization experiment of <a href="https://hal.archives-ouvertes.fr/hal-01167948v1">[1]</a>.
%>
%> The results are stored in "./output/results_BSL_user.mat".
%> DURATION: Computations should take around 3 minutes.
%>
%> The MEG gain matrices used are
%> the precomputed ones in "data/faust_MEG_rcg_X.mat" (in the installation directory of the FAµST toolbox)
%>
%> References:
%>
%> [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
%> approximations of matrices and applications", Journal of Selected
%> Topics in Signal Processing, 2016.
%> <https://hal.archives-ouvertes.fr/hal-01167948v1>
%>
%> [2] A. Gramfort, M. Luessi, E. Larson, D. Engemann, D. Strohmeier,
%> C. Brodbeck, L. Parkkonen, M. Hamalainen, MNE software for processing
%> MEG and EEG data <http://www.ncbi.nlm.nih.gov/pubmed/24161808>,
%> NeuroImage, Volume 86, 1 February 2014, Pages 446-460, ISSN 1053-8119,
%>
%>===============================================================================
function BSL()
import matfaust.Faust
runPath=which(mfilename);
pathname = fileparts(runPath);
% folder_path where the MEG and Faust approximations are stored
BSL_data_pathName=[pathname '..' filesep '..' filesep '..' filesep 'data'];
% useless if the faust toolbox has been properly added in the matlab path
% files where a precomputed FauST approximation of MEG matrix are stored
MEG_FAuST_list_filename={'faust_MEG_rcg_6.mat','faust_MEG_rcg_8.mat','faust_MEG_rcg_16.mat','faust_MEG_rcg_25.mat'};
nb_FAuST_MEG = length(MEG_FAuST_list_filename);
% list of the MEG matrices (the original one and her Faust approximations)
MEG_list = cell(1,nb_FAuST_MEG+1);
nb_MEG_matrix = length(MEG_list);
%% Loading the MEG matrix and the points in the brain
%load([BSL_data_pathName filesep 'matrix_MEG.mat' ]);
load('matrix_MEG.mat')
MEG_matrix = normalizeCol(matrix');% normalization of the columns of the MEG matrix
MEG_list{1}=MEG_matrix;
%% Loading the precomputed FAuST representing the MEG matrix
% RCG (Relative Complexity Gain) : the theoretical speed-up using Faust
RCG_list=zeros(1,nb_FAuST_MEG);
for i=1:nb_FAuST_MEG
%load([BSL_data_pathName filesep MEG_FAuST_list_filename{i}]);
load(MEG_FAuST_list_filename{i});
facts = normalizeCol(facts,lambda); % normalization of the columns of the FAUST
MEG_FAuST=Faust(facts); % construct the FAuST from its factorscons15_row
MEG_list{i+1}=MEG_FAuST; % store the different FAuST approximationscons21_col
RCG_list(i)=rcg(MEG_FAuST); % compute the RCG of the given FAuST
end
M=size(MEG_matrix,2); % Number of points used in the MEG matrix
Ntraining = 500; % Number of training vectorscons22_row
Sparsity = 2; % Number of sources per training vector
dist_paliers = [0.01,0.05,0.08,0.5];
resDist = zeros(nb_MEG_matrix,numel(dist_paliers)-1,Sparsity,Ntraining);% (Matrice,dist_sources,src_nb,run);
compute_Times = zeros(nb_MEG_matrix,numel(dist_paliers)-1,Ntraining);
h = waitbar(0,['Brain Source Localization : MEG matrix and its faust approximations with omp solver']);
nb_palier=numel(dist_paliers)-1;
for k=1:nb_palier;
%Parameters settings
% generates the different source position
Gamma = zeros(M,Ntraining);
for ii=1:Ntraining
dist_sources = -1;
while ~((dist_paliers(k)<dist_sources)&&(dist_sources<dist_paliers(k+1)))
Gamma(:,ii) = sparse_coeffs(MEG_matrix, 1, Sparsity); %
idx = find(Gamma(:,ii));
dist_sources = norm(points(idx(1),:) - points(idx(2),:));
end
end
% compute the data registered by MEG sensor
Data = MEG_matrix*Gamma;
for i=1:Ntraining
waitbar(((k-1)*Ntraining+i)/(nb_palier*Ntraining));
% index of the real source localization
idx = find(Gamma(:,i));
for j=1:nb_MEG_matrix
MEG_FAuST = MEG_list{j};
% find the active source
tic
sol_solver=greed_omp_chol(Data(:,i),MEG_FAuST,M,'stopTol',1*Sparsity,'verbose',false);
compute_Times(j,k,i)=toc;
% compute the distance between the estimated source and the real one
idx_solver = find(sol_solver);
resDist(j,k,1,i) = min(norm(points(idx(1),:) - points(idx_solver(1),:)),norm(points(idx(1),:) - points(idx_solver(2),:)));
resDist(j,k,2,i) = min(norm(points(idx(2),:) - points(idx_solver(1),:)),norm(points(idx(2),:) - points(idx_solver(2),:)));
end
end
end
close(h);
mkdir 'output'
matfile = [ 'output' filesep 'results_BSL_user' ];
save(matfile,'resDist','Sparsity','RCG_list','compute_Times','Ntraining','nb_MEG_matrix');
end
wrapper/matlab/+matfaust/+demo/+bsl/Fig_BSL.m 0 → 100755
+ 215
− 0
View file @ 890a62d5
%% Description Fig_BSL
% Brain source localization figures
% This script builds the BSL figure (Fig. 9.) used in [1].
%
% For more information on the FAuST Project, please visit the website of
% the project : <http://faust.inria.fr>
%
%% License:
% Copyright (2016): Nicolas Bellot, Adrien Leman, Thomas Gautrais, Luc Le Magoarou, Remi Gribonval
% INRIA Rennes, FRANCE
% http://www.inria.fr/
%
% The FAuST Toolbox is distributed under the terms of the GNU Affero
% General Public License.
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU Affero General Public License as published
% by the Free Software Foundation.
%
% This program is distributed in the hope that it will be useful, but
% WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
% See the GNU Affero General Public License for more details.
%
% You should have received a copy of the GNU Affero General Public License
% along with this program. If not, see <http://www.gnu.org/licenses/>.
%
%% Contacts:
% Nicolas Bellot : nicolas.bellot@inria.fr
% Adrien Leman : adrien.leman@inria.fr
% Thomas Gautrais : thomas.gautrais@inria.fr
% Luc Le Magoarou : luc.le-magoarou@inria.fr
% Remi Gribonval : remi.gribonval@inria.fr
%
%% References:
% [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
% approximations of matrices and applications", Journal of Selected
% Topics in Signal Processing, 2016.
% <https://hal.archives-ouvertes.fr/hal-01167948v1>
%
% [2] A. Gramfort, M. Luessi, E. Larson, D. Engemann, D. Strohmeier,
% C. Brodbeck, L. Parkkonen, M. Hamalainen, MNE software for processing
% MEG and EEG data <http://www.ncbi.nlm.nih.gov/pubmed/24161808>,
% NeuroImage, Volume 86, 1 February 2014, Pages 446-460, ISSN 1053-8119,
% [DOI] <http://dx.doi.org/10.1016/j.neuroimage.2013.10.027>
%%
%===============================================================================
%> Brain source localization figures.
%===
%> This script builds the BSL figure (Fig. 9.) used in <a href="https://hal.archives-ouvertes.fr/hal-01167948v1">[1]</a>.
%>
%> References:
%>
%> [1] Le Magoarou L. and Gribonval R., "Flexible multi-layer sparse
%> approximations of matrices and applications", Journal of Selected
%> Topics in Signal Processing, 2016.
%> <https://hal.archives-ouvertes.fr/hal-01167948v1>
%>
%===============================================================================
function Fig_BSL()
runPath=which(mfilename);
pathname = fileparts(runPath);
matfile = ['output' filesep 'results_BSL_user.mat'];
if (not(exist(matfile)))
error('run BSL.m before Fig_BSL.m');
end
%% figure configuration
figure_dir = ['.' filesep 'Figures'];
mkdir(figure_dir)
format_fig='-dpng';
load(matfile);
nb_approx_MEG=nb_MEG_matrix-1;
solver_choice='omp';
Ntest=Ntraining*Sparsity;
%% convergence analysis
d1 = cat(4,resDist(:,1,1,:),resDist(:,1,2,:));
d2 = cat(4,resDist(:,2,1,:),resDist(:,2,2,:));
d3 = cat(4,resDist(:,3,1,:),resDist(:,3,2,:));
test2 = 100*[squeeze(d1);zeros(1,Ntest);squeeze(d2);zeros(1,Ntest);squeeze(d3)];
f=figure('color',[1 1 1]);
T = bplot(test2','linewidth',1.5);
legend(T);
ylabel('Distance between true and estimated sources (cm)')
box on
ax = gca;
x_tick=[1:nb_approx_MEG+1;nb_approx_MEG+3:2*nb_approx_MEG+3;2*nb_approx_MEG+5:3*nb_approx_MEG+5];
mean_x_tick = round(mean(x_tick,2));
set(ax,'XTick',reshape(x_tick',1,numel(x_tick)));
set(ax,'xticklabel', [])
axi = axis;
axis([0 3*(nb_approx_MEG+2) -0.3 4.7])
HorizontalOffset = 10;
verticalOffset = 0.43;
verticalOffset2 = 0.7;
yTicks = get(ax,'ytick');
xTicks = get(ax, 'xtick');
minY = min(yTicks);
text(xTicks(1), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(2+nb_approx_MEG), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(1+2*(nb_approx_MEG+1)), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(1:(nb_approx_MEG+1)))), minY - verticalOffset2, '$1<d<5$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(nb_approx_MEG+2:2*(nb_approx_MEG+1)))), minY - verticalOffset2, '$5<d<8$','HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(round(mean(2*(nb_approx_MEG+1)+1:3*(nb_approx_MEG+1)))), minY - verticalOffset2, '$d>8$','HorizontalAlignment','center','interpreter', 'latex');
for i=1:nb_approx_MEG
text(xTicks(i+1), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(i+2+nb_approx_MEG), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) ' }$' ],'HorizontalAlignment','center','interpreter', 'latex');
text(xTicks(i+1+2*(nb_approx_MEG+1)), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) ' }$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
title(['BSL - convergence (C++ wrapper faust) ' solver_choice ' solver']);
f.Name =['Brain Source Localization : convergence with ' solver_choice '_solver (C++ wrapper)'];
figure_name=[figure_dir filesep 'BSL-convergence_Cpp_' solver_choice '_solver'];
print(figure_name, format_fig);
%% time comparison
timeS = cat(3,compute_Times(:,1,:),compute_Times(:,2,:),compute_Times(:,3,:));
timeS = squeeze(1000*timeS)';
f=figure('color',[1 1 1]);
T = bplot(timeS,'linewidth',1.5);
legend(T);
ylabel('Computed Time (ms)')
box on
ax = gca;
x_tick=[1:2*nb_approx_MEG+1];
mean_x_tick = round(mean(x_tick,2));
set(ax,'XTick',x_tick');
set(ax,'xticklabel', [])
axi = axis;
HorizontalOffset = 10;
verticalOffset = 0.5;
verticalOffset2 = 1;
yTicks = get(ax,'ytick');
xTicks = get(ax, 'xtick');
minY = min(yTicks);
text(xTicks(1), minY - verticalOffset, '$\mathbf{M}$','HorizontalAlignment','center','interpreter', 'latex');
for i=1:nb_approx_MEG
text(xTicks(i+1), minY - verticalOffset, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
title(['BSL - time comparison (FAUST vs dense matrix) ' solver_choice ' solver']);
f.Name =['Brain Source Localization : time comparison with ' solver_choice 'solver'];
figure_name=[figure_dir filesep 'BSL-time_comparison_' solver_choice '_solver'];
print(figure_name, format_fig);
%% speed-up
mean_Times=mean(timeS);
dense_matrix_time=mean_Times(1);
real_RCG=dense_matrix_time./mean_Times;
f=figure;
ax = gca;
set(ax,'xticklabel', [])
set(ax,'Visible','off');
plot(1:nb_approx_MEG,real_RCG(2:end),'linewidth',1.5);
hold on
plot(1:nb_approx_MEG,ones(1,nb_approx_MEG),'linewidth',1.5);
verticalOffset=1.5;
minY=min([0.9,real_RCG(2:end)]);
maxY=max([0.9,real_RCG(2:end)]);
axis([1,nb_approx_MEG,minY,maxY]);
set(ax,'XTick',[]);
for i=1:nb_approx_MEG
text(i, minY -(maxY-minY)/20, ['$\widehat{\mathbf{M}}_{' int2str(RCG_list(i)) '}$' ],'HorizontalAlignment','center','interpreter', 'latex');
end
legend('speed up FAuST','neutral speed up');
title(['BSL - speed up using FAUST ' solver_choice ' solver']);
f.Name =['Brain Source Localization : speed-up Faust with ' solver_choice 'solver'];
figure_name = [figure_dir filesep 'BSL-speed_up_' solver_choice ' solver'];
print(figure_name, format_fig);
%% console display
disp(['**** MEG with OMP solver time comparison ****']);
disp(['M tps : ' num2str(mean(timeS(:,1))) ]);
for i=1:nb_approx_MEG
disp([ 'M_' int2str(RCG_list(i)) ' tps : ' num2str(mean(timeS(:,i+1))) ', speed-up : ' num2str(real_RCG(i+1))]);
end
end
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment