diff --git a/parmorceauxMax-git/HTN319.txt b/parmorceauxMax-git/HTN319.txt
new file mode 100644
index 0000000000000000000000000000000000000000..786e68a6ed4f9b903c503df89a9cd545a9525295
--- /dev/null
+++ b/parmorceauxMax-git/HTN319.txt
@@ -0,0 +1,25 @@
+1.926700000000000000e+04
+1.902000000000000000e+04
+1.874400000000000000e+04
+1.815700000000000000e+04
+1.702000000000000000e+04
+1.368000000000000000e+04
+1.148600000000000000e+04
+9.703000000000000000e+03
+8.250000000000000000e+03
+7.261000000000000000e+03
+6.785000000000000000e+03
+6.931000000000000000e+03
+7.017000000000000000e+03
+7.391000000000000000e+03
+8.095000000000000000e+03
+9.878000000000000000e+03
+1.243700000000000000e+04
+1.421700000000000000e+04
+1.539900000000000000e+04
+1.631200000000000000e+04
+1.698000000000000000e+04
+1.761900000000000000e+04
+1.825400000000000000e+04
+1.882700000000000000e+04
+1.920400000000000000e+04
diff --git a/parmorceauxMax-git/HTN320.txt b/parmorceauxMax-git/HTN320.txt
new file mode 100644
index 0000000000000000000000000000000000000000..dbaf22844fb82d2112a66d6ef61e2986c77bd8d7
--- /dev/null
+++ b/parmorceauxMax-git/HTN320.txt
@@ -0,0 +1,25 @@
+2.163700000000000000e+04
+1.804500000000000000e+04
+1.839600000000000000e+04
+1.878900000000000000e+04
+1.517800000000000000e+04
+1.375100000000000000e+04
+1.200000000000000000e+04
+1.040900000000000000e+04
+9.332000000000000000e+03
+9.682000000000000000e+03
+9.296000000000000000e+03
+9.940000000000000000e+03
+1.015600000000000000e+04
+1.024000000000000000e+04
+1.074500000000000000e+04
+1.157100000000000000e+04
+1.227300000000000000e+04
+1.375700000000000000e+04
+1.443800000000000000e+04
+1.507700000000000000e+04
+1.559700000000000000e+04
+1.641400000000000000e+04
+1.733300000000000000e+04
+1.793800000000000000e+04
+1.841100000000000000e+04
diff --git a/parmorceauxMax-git/HTN321.txt b/parmorceauxMax-git/HTN321.txt
new file mode 100644
index 0000000000000000000000000000000000000000..3c4d3f1add965dd05c2d71219693a9f233f81c48
--- /dev/null
+++ b/parmorceauxMax-git/HTN321.txt
@@ -0,0 +1,25 @@
+1.954400000000000000e+04
+1.934700000000000000e+04
+1.794400000000000000e+04
+1.577000000000000000e+04
+1.431800000000000000e+04
+1.279900000000000000e+04
+1.091300000000000000e+04
+9.179000000000000000e+03
+7.974000000000000000e+03
+7.012000000000000000e+03
+6.230000000000000000e+03
+5.746000000000000000e+03
+5.827000000000000000e+03
+6.672000000000000000e+03
+7.861000000000000000e+03
+1.021100000000000000e+04
+1.172300000000000000e+04
+1.356200000000000000e+04
+1.471400000000000000e+04
+1.574300000000000000e+04
+1.641500000000000000e+04
+1.699600000000000000e+04
+1.748800000000000000e+04
+1.842300000000000000e+04
+1.881800000000000000e+04
diff --git a/parmorceauxMax-git/HTN322.txt b/parmorceauxMax-git/HTN322.txt
new file mode 100644
index 0000000000000000000000000000000000000000..b0380c53fb117c6d61f985db719fe32305bb9ed0
--- /dev/null
+++ b/parmorceauxMax-git/HTN322.txt
@@ -0,0 +1,25 @@
+2.706000000000000000e+04
+2.273200000000000000e+04
+2.073500000000000000e+04
+1.713100000000000000e+04
+1.408500000000000000e+04
+1.138800000000000000e+04
+9.589000000000000000e+03
+8.362000000000000000e+03
+7.824000000000000000e+03
+8.599000000000000000e+03
+7.907000000000000000e+03
+8.832000000000000000e+03
+9.581000000000000000e+03
+1.205800000000000000e+04
+1.363100000000000000e+04
+1.511600000000000000e+04
+1.646000000000000000e+04
+1.784500000000000000e+04
+1.958700000000000000e+04
+2.142300000000000000e+04
+2.266700000000000000e+04
+2.374000000000000000e+04
+2.445000000000000000e+04
+2.505300000000000000e+04
+2.541000000000000000e+04
diff --git a/parmorceauxMax-git/Interface.asv b/parmorceauxMax-git/Interface.asv
new file mode 100644
index 0000000000000000000000000000000000000000..5ea2b494ecd802460b69d680da51325c1f0d3a42
--- /dev/null
+++ b/parmorceauxMax-git/Interface.asv
@@ -0,0 +1,377 @@
+classdef Interface < handle
+ properties
+ %% données d'entrée
+ T %temps (ou numéro de frame)
+ Y %observation : volume du ventricule gauche. Une observation (frame) par ligne de Y
+ N
+
+ %% résultats en sortie
+ meilleure_sequence
+ meilleure_sequence_models
+ loglikelihood
+
+ %% modèle statistique
+ interpolant_filenames
+
+ %% morceaux courants (cad non finis) A, B1, B2 et C
+ dynamics_models
+ first_frame_of_models
+
+ %% paramètres de la méthode
+ initial_alphas
+
+ sigma_estimation
+ sigma_test
+
+ threshold_dynamics
+
+ display_screen
+
+ %% Viterbi
+ transition
+
+ proba_morceaux % probabilité courante pour chaque morceau, avec son meilleur path Viterbi
+ morceaux_finis % pour chaque morceau courant : liste des morceaux finis du path. Cette liste contient une copie du morceau dans l'état où il était quand il a été terminé
+ proba_morceaux_finis % pour chaque morceau courant : liste des probabilités associées aux morceaux finis. Elle combine la proba de chaque observation du morceau et la proba de transition du morceau au suivant
+
+ model_indexes % pour chaque morceau courant : liste des indices des morceaux du path Viterbi pour chaque frame
+ scores_finis % pour chaque morceau courant : liste des proba d'observation pour chaque frame (uniquement les frames des morceaux finis, car le morceau courant n'est pas encore stable et les x peuvent encore changer)
+
+ end
+ methods
+ function obj = Interface(interpolant_filenames, display_screen)
+
+ obj.interpolant_filenames = interpolant_filenames;
+
+% obj.transition = [ 0.5 0.5 0 0; % départ de A
+% 0 1/3 1/3 1/3; % départ de B1
+% 0 0 0.5 0.5; % départ de B2
+% 0 0 0 1 % départ de C
+% ];
+
+ tr = load("trans.mat");
+ obj.transition = tr.trans;
+ obj.display_screen = display_screen;
+ end
+ function process_new_sequence(obj, filename, initial_alphas, sigma_estimation, sigma_test, threshold_dynamics)
+ obj.threshold_dynamics = threshold_dynamics;
+
+ obj.dynamics_models = {NaN, NaN, NaN, NaN}; % modèles courants A B1 B2 et C
+ obj.first_frame_of_models = [1 NaN NaN NaN]; % frame où chaque modèle commence
+
+ if obj.display_screen == -1
+ obj.dynamics_models{1} = Dynamics_Model(obj.interpolant_filenames{1}, obj.display_screen);
+ else
+ obj.dynamics_models{1} = Dynamics_Model(obj.interpolant_filenames{1}, obj.display_screen+1);
+ end
+
+ obj.proba_morceaux = [1 0 0 0]; % proba des modèles A, B1, B2, et C
+
+ obj.T = [];
+ obj.Y = [];
+ obj.N = [];
+
+ obj.scores_finis = {[], [], [], []};
+ obj.model_indexes = {[], [], [], []};
+
+ obj.initial_alphas = initial_alphas;
+ obj.sigma_estimation = sigma_estimation;
+ obj.sigma_test = sigma_test;
+
+ obj.morceaux_finis = {[], [], [], []};
+ obj.proba_morceaux_finis = {[], [], [], []};
+
+ obj.read_data(filename);
+ obj.N = size(obj.T,1);
+
+ obj.dynamics_models{1}.initialise(0, initial_alphas(1), sigma_estimation, sigma_test,1);
+ %ligne 14
+ for frame=1:obj.N
+ obj.process_frame(frame);
+
+ if obj.display_screen ~= -1
+ for model=1:4
+ if obj.proba_morceaux(model) == 0 || isnan(obj.proba_morceaux(model))
+ continue
+ end
+ obj.dynamics_models{model}.plot_estimated_X();
+ end
+ end
+ end
+
+
+ %% choix de la meilleure sequence (peut être pour plus tard: toujours choisir la séquence qui se termine sur le modèle C ?)
+ [max_proba, indice] = max(obj.proba_morceaux);
+ obj.meilleure_sequence = obj.model_indexes{indice};
+ obj.meilleure_sequence_models = [obj.morceaux_finis{indice} obj.dynamics_models{indice}];
+
+ %% calcul les scores pour les frames du dernier morceau
+ % mise à jour des x pour bien les étaler
+ obj.meilleure_sequence_models(end).spread_X();
+ new_scores = obj.meilleure_sequence_models(end).compute_loglikelihood_dynamics([1:obj.N-obj.first_frame_of_models(indice)+1]);
+ obj.loglikelihood = [obj.scores_finis{indice} new_scores];
+
+ %% à faire : afficher la séquence des modèles et la séquence de scores comme dans plot
+ close all
+ for i=1:length(obj.meilleure_sequence_models)
+ obj.meilleure_sequence_models(i).plot_estimated_X();
+ end
+%%% save plot
+
+ end
+ function read_data(obj, filename)
+ obj.Y =load(filename);
+ %obj.T = [1:75]';
+ %obj.Y = obj.Y / max(obj.Y);
+ %obj.Y = repmat(obj.Y, 3, 1);
+ obj.T = [1:25]';
+ obj.Y = obj.Y / max(obj.Y);
+ end
+ function process_frame(obj, frame_number)
+
+ frame_number
+
+ Yt = obj.Y(frame_number,:);
+ Tt = obj.T(frame_number);
+
+ %% tableaux temporaires dans lesquels on prépare la prochaine version des tableaux de la classe sans les modifier avant d'avoir terminé de traiter la frame pour tous les morceaux
+ new_proba_morceaux = [NaN NaN NaN NaN];
+ new_morceaux_finis = {[], [], [], []};
+ new_proba_morceaux_finis = {[], [], [], []};
+
+ new_models = {NaN, NaN, NaN, NaN};
+ new_first_frame_of_models = [NaN NaN NaN NaN];
+
+ new_scores_finis = {[], [], [], []};
+ new_model_indexes = {[], [], [], []};
+
+
+ %% boucle sur les 4 modèles possibles
+ for morceau = 1:4
+
+ % on va regarder toutes les branches qui arrivent sur le modèle
+
+ %% tableaux temporaires pour le morceau courant et toutes les branches possibles
+ modele_branche = {NaN, NaN, NaN, NaN}; % copie des modèles pour leur montrer la nouvelle frame, sans modifier ceux de obj.dynamics_models
+ first_frame_of_model_branche = [NaN NaN NaN NaN]; % pour les copies des modèles
+ proba_obs_branche = [NaN NaN NaN NaN]; % proba d'observation donnée par les copies des modèles ayant vu la frame courante
+ nouveau_morceau_fini_branche = {NaN, NaN, NaN, NaN}; % si une branche implique un nouveau morceau fini, on le sauvegarde ici en attendant Viterbi et le choix de la branche (cad si on le garde ou pas)
+ proba_branche = [NaN NaN NaN NaN]; % proba totale de chaque branche
+ proba_morceau_prec = [NaN NaN NaN NaN]; % proba totale des frames vues par le morceau qui était courant à la frame précédente (donc d'indice=branche). Ces proba sont combinées avec les probas de transition au sein du morceau
+ new_scores_branche = {[] [] [] []}; % score des frames vues par le morceau qui était courant à la frame précédente
+
+ if frame_number == 1
+ if obj.proba_morceaux(morceau) == 0
+ continue
+ end
+
+ new_models{morceau} = copy(obj.dynamics_models{morceau});
+ new_first_frame_of_models(morceau) = obj.first_frame_of_models(morceau);
+
+ new_models{morceau}.process_new_frame(Yt, Tt);
+ proba_obs = new_models{morceau}.compute_loglikelihood_dynamics(1);
+
+ new_proba_morceaux(morceau) = log(obj.proba_morceaux(morceau)) + proba_obs;
+ new_morceaux_finis{morceau} = [];
+
+ new_model_indexes{morceau} = [morceau];
+ else
+ %% on calcule les probas de toutes les branches qui arrivent sur le morceau courant
+ for branche = 1:4
+ duree=1;
+ %trans = obj.transition{frame_number}(branche, morceau);
+ trans = obj.transition{branche, morceau}(frame_number);
+ if frame_number>2
+ duree=sum(obj.model_indexes{branche}==branche);
+ if (duree>0)
+ %trans = obj.transition{duree}(branche, morceau);
+ trans = obj.transition{branche, morceau}(duree);
+ end
+ end
+
+
+ % on commence par ignorer les branches impossibles
+ if trans == 0
+ continue
+ end
+ if obj.proba_morceaux(branche) == 0 || isnan(obj.proba_morceaux(branche))
+ continue
+ end
+
+ if branche == morceau % cas sans changement de modèle, donc le modèle est déjà initialisé
+ modele_branche{branche} = copy(obj.dynamics_models{branche});
+ first_frame_of_model_branche(branche) = obj.first_frame_of_models(morceau);
+ else % cas avec changement de modèle, donc il faut créer et initialiser le modèle
+ %% on crée le modèle et on l'initialise
+ if obj.display_screen == -1
+ modele_branche{branche} = Dynamics_Model(obj.interpolant_filenames{morceau}, obj.display_screen);
+ else
+ modele_branche{branche} = Dynamics_Model(obj.interpolant_filenames{morceau}, obj.display_screen+morceau);
+ end
+ modele_branche{branche}.initialise(0, obj.initial_alphas(morceau), obj.sigma_estimation, obj.sigma_test,morceau);
+ %modele_branche{branche}.initialise(0, obj.dynamics_models{branche}.current_alpha, obj.sigma_estimation, obj.sigma_test);
+ first_frame_of_model_branche(branche) = frame_number;
+ end
+
+ %% on donne la frame au modèle, et on récupère la proba d'observation (dans la fenêtre)
+ modele_branche{branche}.process_new_frame(Yt, Tt);
+ %print branche et morceau
+ BRANCHE=branche
+ MORCEAU=morceau
+ proba_obs_branche(branche) = modele_branche{branche}.compute_loglikelihood_dynamics(frame_number-first_frame_of_model_branche(branche)+1);
+
+ %% on garde en mémoire les morceaux qui sont finis, pour l'affichage final des X
+%%%% %%%%%%%%%%%%%%%% a faire : mettre à jour les x pour bien les étaler
+ if branche == morceau
+ nouveau_morceau_fini_branche{branche} = 0;
+ else
+ nouveau_morceau_fini_branche{branche} = copy(obj.dynamics_models{branche}); % ce modèle n'a pas encore vu la frame courante, on peut donc le sauvegader dans son état à la frame précédente
+%%%% %%%%%% mettre à jour la copie ici
+ nouveau_morceau_fini_branche{branche}.spread_X();
+ end
+
+ %% on récupère les scores des frames vues par le modèle
+ if branche == morceau
+ %% on récupère les scores pour les frames précédentes, qui ont pu bouger avec le argmax
+ proba_morceau_prec(branche) = sum(modele_branche{branche}.compute_loglikelihood_dynamics([1:frame_number-first_frame_of_model_branche(branche)]));
+ % proba_morceau_prec(branche) = proba_morceau_prec(branche) + log(trans) * (frame_number-first_frame_of_model_branche(branche)-1);
+ %test sum(obj.transition{1,2}(9:10))
+
+ transi=sum(obj.transition{branche,branche}(1:frame_number-first_frame_of_model_branche(branche)-1));
+ proba_morceau_prec(branche) = proba_morceau_prec(branche) + log(transi);
+ else
+ %% on récupère les scores pour les frames précédentes, qui ont pu bouger avec le argmax
+ new_scores_branche{branche} = nouveau_morceau_fini_branche{branche}.compute_loglikelihood_dynamics([1:frame_number-obj.first_frame_of_models(branche)]);
+ proba_morceau_prec(branche) = sum(new_scores_branche{branche});
+ %trans_morceau_prec = obj.transition{duree}(branche, branche);
+ %proba_morceau_prec(branche) = proba_morceau_prec(branche) + log(trans_morceau_prec) * (frame_number-obj.first_frame_of_models(branche)-1);
+ trans_morceau_prec = sum(obj.transition{branche,morceau}(1:frame_number-obj.first_frame_of_models(branche)))
+ proba_morceau_prec(branche) = proba_morceau_prec(branche) + log(trans_morceau_prec);
+ end
+
+ proba_branche(branche) = sum(obj.proba_morceaux_finis{branche}) + proba_morceau_prec(branche) + log(trans) + proba_obs_branche(branche);
+ %duree
+ end
+
+ %% on choisi la meilleure branche
+ [max_val, indice] = max(proba_branche);
+
+ if isnan(max_val)
+ continue
+ end
+
+ new_proba_morceaux(morceau) = max_val;
+
+ new_models{morceau} = modele_branche{indice};
+ new_first_frame_of_models(morceau) = first_frame_of_model_branche(indice);
+
+ if nouveau_morceau_fini_branche{indice} ~= 0
+ new_morceaux_finis{morceau} = [obj.morceaux_finis{indice} nouveau_morceau_fini_branche{indice}];
+ %trans=obj.transition{frame_number}(indice,morceau);
+ trans=obj.transition{indice,morceau}(frame_number);
+ if duree>0
+ %trans = obj.transition{duree}(indice, morceau);
+ trans = obj.transition{indice, morceau}(duree);
+ end
+ new_proba_morceaux_finis{morceau} = [obj.proba_morceaux_finis{indice} proba_morceau_prec(indice) + log(trans)];
+ new_scores_finis{morceau} = [obj.scores_finis{indice} new_scores_branche{indice}];
+ else
+ new_morceaux_finis{morceau} = obj.morceaux_finis{indice};
+ new_proba_morceaux_finis{morceau} = obj.proba_morceaux_finis{indice};
+ new_scores_finis{morceau} = obj.scores_finis{indice};
+ end
+
+ new_model_indexes{morceau} = [obj.model_indexes{indice} morceau];
+ end
+ end
+
+ obj.proba_morceaux = new_proba_morceaux;
+ obj.dynamics_models = new_models;
+ obj.first_frame_of_models = new_first_frame_of_models;
+
+ obj.scores_finis = new_scores_finis;
+
+ obj.morceaux_finis = new_morceaux_finis;
+ obj.proba_morceaux_finis = new_proba_morceaux_finis;
+
+ obj.model_indexes = new_model_indexes;
+
+
+ %% réflexions pour plus tard
+ %% gérer : alpha différents pour les différents modèles (alpha_ini + marges)
+ %% fin des modèles : duppliquer les dernières colonnes pour que x ait le temps de se fixer ? (avec pénalité lors de la comparaison des modèles)
+ %% ou alors : on n'attend pas que les x soient stables
+ %% initialiser le modèle concurrent à chaque fois (tant qu'on n'a pas basculé dessus)
+ end
+ function plot(obj) %% cette fonction est à modifier car maintenant nous avons plusieurs modèles / morceaux
+ obj.dynamics_model.plot_estimated_X();
+
+ figure(obj.display_screen);
+ n = length(obj.loglikelihood_dynamics_window);
+ subplot(4,1,1)
+ hold off
+ plot(obj.T(1:n),obj.Y(1:n,1))
+ title('Pose representation')
+ xlabel('frame number')
+ ylabel('1st dimension')
+ XL = xlim;
+
+ subplot(4,1,2)
+ hold off
+ plot(obj.T(1:n),obj.loglikelihood_pose)
+ hold on
+ indexes_p = (obj.loglikelihood_pose < obj.threshold_pose);
+ plot(obj.T(indexes_p),obj.loglikelihood_pose(indexes_p),'.r')
+ title('Likelihood of the pose to be normal')
+ xlabel('Frame number')
+ ylabel('Log likelihood')
+ xlim(XL)
+ YL = ylim;
+ YL1 = max(YL(1),-5);
+ ylim([YL1 YL(2)])
+ hold on
+
+ subplot(4,1,3)
+ hold off
+ plot(obj.T(1:n),obj.loglikelihood_dynamics_window)
+ hold on
+ indexes_w = (obj.loglikelihood_dynamics_window < obj.threshold_dynamics);
+ plot(obj.T(indexes_w),obj.loglikelihood_dynamics_window(indexes_w),'.r')
+ title('Likelihood of the dynamics to be normal, computed in the dynamic window')
+ xlabel('Frame number')
+ ylabel('Log likelihood')
+ xlim(XL)
+ YL = ylim;
+ YL1 = max(YL(1),-20);
+ ylim([YL1 YL(2)])
+ hold on
+
+ subplot(4,1,4)
+ hold off
+ n2 = length(obj.loglikelihood_dynamics);
+ plot(obj.T(1:n2),obj.loglikelihood_dynamics)
+ hold on
+ indexes_d = (obj.loglikelihood_dynamics < obj.threshold_dynamics);
+ plot(obj.T(indexes_d),obj.loglikelihood_dynamics(indexes_d),'.r')
+ title('Likelihood of the dynamics to be normal, computed for each converged frame')
+ xlabel('Frame number')
+ ylabel('Log likelihood')
+ xlim(XL)
+ YL = ylim;
+ YL1 = max(YL(1),-20);
+ YL2 = min(YL(2),10);
+ ylim([YL1 YL2])
+ hold on
+
+ subplot(4,1,1)
+ hold on
+ plot(obj.T(indexes_p),obj.Y(indexes_p,1),'.r')
+ %plot(obj.T(indexes_w),obj.Y(indexes_w,1),'.r')
+ plot(obj.T(indexes_d),obj.Y(indexes_d,1),'.r')
+ %indexes = indexes_p == 0 & indexes_w == 0 & [indexes_d ones(1,n-n2)] == 0;
+ indexes = indexes_p == 0 & [indexes_d ones(1,n-n2)] == 0;
+ plot(obj.T(indexes),obj.Y(indexes,1),'.g')
+ end
+
+ end
+end
diff --git a/parmorceauxMax-git/ROC.m b/parmorceauxMax-git/ROC.m
index c2391a4d2488cb2d9126a95591d2e463a9a04209..0bf1c824754a7646056511d64d98f1f67acc5afc 100644
--- a/parmorceauxMax-git/ROC.m
+++ b/parmorceauxMax-git/ROC.m
@@ -1,6 +1,9 @@
function [F_mesure,M_seuil]=ROC()
- Seuil=[-9:0.1:1];
- diagn=[true,false,true,false,false,true,false,false,false,false,true,false,true,false,false,true,false,false,false,false,true,false,false,true,true,true,true,false,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true];
+ %Seuil=[-9:0.1:1];
+ %Seuil=[0:0.001:5];%[-6:0.01:5];%[-6:0.001:5];%[-3:0.001:7]
+ Seuil=[-5:0.01:6];
+ diagn=[true,false,false,false,true,false,false,false,false,true,false,true,false,false,false,false,false,false,false,false,true,true,true,false,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true];
+ %[true,false,true,false,false,true,false,false,false,false,true,false,true,false,false,true,false,false,false,false,true,false,false,true,true,true,true,false,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true]
a=1;
listept=dir(['LLLDyn*.txt']);
X=zeros(1,47);
@@ -9,7 +12,7 @@ function [F_mesure,M_seuil]=ROC()
R=zeros(1,47);
dist1=2^(1/2);
F_mesure1=0;
- while Seuil(a)~=1
+ while Seuil(a)~=Seuil(end)
seuil=Seuil(a);
[TN,TP,FN,FP]=TN_TP_FN_FP(listept,seuil,diagn);
TPR=TP/(TP+FN);
@@ -39,10 +42,11 @@ function [F_mesure,M_seuil]=ROC()
dist2=(FPR^2+(TPR-1)^2)^(1/2)
M_seuil
%Seuil(81);
- plot(X,Y);
+ plot(X,Y)
+ %hold on
+ %plot([0 1],[0 1])
%(1,82);
%X(1,82);
-
[X2,I]=sort(X);
Y2=Y(I);
len=length(X2);
diff --git a/parmorceauxMax-git/example_test_models.asv b/parmorceauxMax-git/example_test_models.asv
new file mode 100644
index 0000000000000000000000000000000000000000..e16010f4d8cac7c4e79e15132ef08bc4cdb021b7
--- /dev/null
+++ b/parmorceauxMax-git/example_test_models.asv
@@ -0,0 +1,195 @@
+%F_Mesure=0;
+%M_Seuil=0;
+%for initial_alpha=0.035:0.001:0.043%0.035:0.001:0.045%0.01:0.01:0.1%0.05:0.05
+ %% liste of testing data
+ liste = dir(['HTN*.txt']);
+ save testtest1.mat
+ save testtest2.mat
+ save LLLDynWin.mat
+ save LLLDyn.mat
+ %% initialisation
+ % set display to -1 to disable
+ display = -1;%1(ancienne version)%7;%-1
+ interface = Interface({'interpolants_A', 'interpolants_B1', 'interpolants_B2', 'interpolants_C'}, display);
+ % set initial_alpha = -1 to have it estimated automatically
+ initial_alpha =0.037% 0.038 %0.04;%0.05 %0.025;(0.039)
+ sigma_estimation = 7e-3;
+ sigma_test = 1e-3;
+ threshold_dynamics = -0.7;
+ %% testing
+ J=[9];%[2,3,4,7,8,9,11,13,14,16,17,18,19,20,24,25,33,39,43,45,47];%J= [8,9,10,14,15,16,20,22,23,27,28,29,33,37,38,42,43]%2,4,5,%J=[2,4,5,8,9,10,11,14,15,20,22,27,28,29,43,46,47]%[8,29,42,45,46]%J=[1,3,6,11,13,16,21,24,25,26,27,29];
+% patient_prbl_1=[];
+% for i=1:47
+% for j=1:25
+% if (LLLDyn(i,j)<=3.7910)
+% patient_prbl_1(end+1)=i;
+% break
+% end
+% end
+% end
+% patient_prbl_1
+% listept=dir(['LLLDynpt*.txt']);
+% diagn=[true,false,false,false,true,false,false,false,false,true,false,true,false,false,false,false,false,false,false,false,true,true,true,false,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true];
+% X=[0:1/24:1];
+% Y=zeros(1,25);
+% cpt=0
+% patientundred=[]
+% for i=1:length(listept)
+% suppr=false;
+% x=load(listept(i).name)
+% % for j=1:25
+% % j
+% % if x(j)<=3.5750;
+% % suppr=true;
+% % patientundred(end+1)=i;
+% % break
+% % end
+% % end
+% if (suppr==false)
+% i
+% if (diagn(i)==true)
+% hold on
+% plot(x,'Color','#ffaa45')
+% else
+% hold on
+% plot(x,'Color','#6da0d6','tag',listept(i).name)
+%
+% end
+%
+% else
+% continue
+% end
+% end
+% hold on
+% yline(0.5510)
+% patientundred
+% datacursormode on
+% dcm = datacursormode(gcf);
+% set(dcm,'UpdateFcn',@myupdatefcn)
+% function txt = myupdatefcn(trash,event)
+% pos = get(event,'Position');
+% dts = get(event.Target,'Tag');
+% txt = {dts,...
+% ['X: ',num2str(pos(1))],...
+% ['Y: ',num2str(pos(2))]};
+ for i=1:length(J)%length(J)%length(liste)
+ filename=strcat(liste(J(i)).name)%liste(i).name%liste(J(i)).name
+ interface.process_new_sequence(filename,[0.1059 0.1622 0.2712 0.45] , sigma_estimation, sigma_test, threshold_dynamics);%[1/(11.6-1) 1/(7.1-1) 1/(4.9-1) 1/(3.2-1)] [1/(10.667-1) 1/(7.278-1) 1/(4.929-1) 1/(3.222-1)][0.1034 0.1651 0.2586 0.45]
+ %j'essaie de "save" les meilleurs sequences
+ testtest1=interface.meilleure_sequence
+ testtest2=interface.loglikelihood
+ if (i==1)
+ bestSequence=testtest1;
+ LLLDyn=testtest2;
+ end
+ if (i>1)
+ bestSequence=[bestSequence;testtest1];
+ LLLDyn=[LLLDyn;testtest2];
+ end
+
+ %LLLDyn=interface.loglikelihood
+
+%% détection de patient avec problème
+%
+% for j=LLLDyn
+% if (j<4.5530)
+% patient_prbl_1(end+1)=i;
+% break
+% end
+% end
+
+%% détection patient avec plateau
+% cpt=0;
+% for j=testtest1
+% if (j==3)
+% cpt=cpt+1;
+% if (cpt==3)
+% patient_plateau(end+1)=i;
+% break
+% end
+% end
+% end
+
+%% save des scores
+
+% testtest1=interface.loglikelihood_dynamics_window(26:50);
+% testtest2=interface.loglikelihood_dynamics(26:50);
+% if (i==1)
+% LLLDynWin=testtest1;
+% LLLDyn=testtest2;
+% end
+% if (i>1)
+% LLLDynWin=[LLLDynWin;testtest1];
+% LLLDyn=[LLLDyn;testtest2];
+% end
+
+%% on regroupe les 4 images en une seule figure
+
+% onion = imread('ModèleA.png');
+% onionArray = repmat(onion, [ 1 1 1 4 ]);
+% im=montage(onionArray,'size',[1 NaN])
+% saveas(im, 'final_model.png')
+%
+% model=["ModèleA.png" "ModèleB1.png" "ModèleB2.png" "ModèleC.png"];
+% for w=1:3
+% img=imread(model(w));
+% subplot(1,3,w)
+% imshow(img);
+% end
+% J(i)
+ % liste(J(i)).name
+ %i
+ end
+
+ %% affichage
+
+% patient_prbl_1
+% patient_plateau
+ %figure()
+ %plot([1:25],LLLDynWin)
+ %figure()
+ %plot([1:25],LLLDyn)
+ %
+
+ for i=1:length(liste)%length(J)%length(liste)
+ if (i < 10)
+ name = strcat('0',int2str(i));
+ else
+ name = int2str(i);
+ end
+
+ dlmwrite(strcat('LLLDynpt',name,'.txt'),LLLDyn(i,:));
+ end
+% %% plot score avec couleurs pour distinguer patients sains
+% listept=dir(['LLLDyn*.txt']);
+% diagn=[true,false,false,false,true,false,false,false,false,true,false,true,false,false,false,false,false,false,false,false,true,true,true,false,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true,true];
+% for i=1:length(diagn)
+% hold on
+% if (diagn==true)
+% plot(LLLDyn(i),'Color','#ffaa45')
+% else
+% plot(LLLDyn(i))
+% end
+% end
+% plot(LLLDyn)
+ %[f_mesure,m_seuil]=find_seuil_score()
+ %if f_mesure>F_Mesure
+ %F_Mesure=f_mesure
+ %optinitial_alpha=initial_alpha
+ %M_Seuil=m_seuil
+ %end
+%end
+%F_Mesure
+%M_Seuil
+%optinitial_alpha
+
+%% code pour extraire les patients détectés malade
+% for i=1:47
+% for j=1:25
+% if (LLLDyn(i,j)<3.9410)
+% patient_prbl_1(end+1)=i;
+% break
+% end
+% end
+% end
+% patient_prbl_1
diff --git a/parmorceauxMax-git/interpolants_A.mat b/parmorceauxMax-git/interpolants_A.mat
index ba6c4024b3794859db242eeaa5b466c0cb3e3f35..f2b9174e951683e3daa0a560b31a7f63d07ec3fa 100644
Binary files a/parmorceauxMax-git/interpolants_A.mat and b/parmorceauxMax-git/interpolants_A.mat differ
diff --git a/parmorceauxMax-git/interpolants_B1.mat b/parmorceauxMax-git/interpolants_B1.mat
index e6b579fc69c10bd0535ec2ac3899294557ba2e26..598a280a3a434ebff8b31f257cdec16e4fae40f3 100644
Binary files a/parmorceauxMax-git/interpolants_B1.mat and b/parmorceauxMax-git/interpolants_B1.mat differ
diff --git a/parmorceauxMax-git/interpolants_B2.mat b/parmorceauxMax-git/interpolants_B2.mat
index 3cd68f84be0f38631c3ab5d8f3365a4679e92591..f797fb1a8c054ad7aaaca388f7a022e5590107d8 100644
Binary files a/parmorceauxMax-git/interpolants_B2.mat and b/parmorceauxMax-git/interpolants_B2.mat differ
diff --git a/parmorceauxMax-git/interpolants_C.mat b/parmorceauxMax-git/interpolants_C.mat
index d8aa33fb9da793cd1b8918dbe550d6b44586d8f1..cc5b524f9ce1de4e071ecac59eaef729148c0ec8 100644
Binary files a/parmorceauxMax-git/interpolants_C.mat and b/parmorceauxMax-git/interpolants_C.mat differ
diff --git a/parmorceauxMax-git/matrice_transition.asv b/parmorceauxMax-git/matrice_transition.asv
new file mode 100644
index 0000000000000000000000000000000000000000..aacbb0d07e65225523dcac51e55d869ae8f9757a
--- /dev/null
+++ b/parmorceauxMax-git/matrice_transition.asv
@@ -0,0 +1,192 @@
+function transition=matrice_transition(eps)
+%C=xlsread('p_t_test.xls');
+C=xlsread('proba_transition.xls');
+%transition=repmat({[zeros(3,4);[0 0 0 1]]},1,sum(C(1,:)));
+N=sum(C(1,:));
+A=repmat({zeros(1,N)}, 3, 4);
+B=repmat({zeros(1,N)}, 1, 3);
+transition=[A; [B, {ones(1,N)}]];
+for i=[1,3]
+ M=max(C(:,i));
+ for j=1:M
+ ind_col=C(:,i)>j
+ ind_col_prec=C(:,i)>=j
+ num1=sum(ind_col)
+ num2=(sum(ind_col_prec)-sum(ind_col))
+ if sum(ind_col)==0
+ transition{i,i}(j)=eps;
+ transition{i,i+1}(j)=num2/sum(ind_col_prec)-eps
+ elseif (sum(ind_col_prec)-sum(ind_col))==0
+ transition{i,i}(j)=num1/sum(ind_col_prec)-eps
+ transition{i,i+1}(j)=eps;
+ else
+ transition{i,i}(j)=num1/sum(ind_col_prec);
+ transition{i,i+1}(j)=num2/sum(ind_col_prec);
+ end
+ end
+ for k=M:sum(C(1,:))
+ transition{i,:}(k)=transition{}(M);
+ end
+end
+for i=2
+ M=max(C(:,i));
+ for j=1:M
+ ind_col=C(:,i)>j;
+ ind_col_prec=C(:,i)>=j;
+ ind_colb1=C(:,2)>j;
+ ind_colc=C(:,i)==j;
+ if sum(ind_colb1)==0
+ if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+ transition{j}(i,i)=eps;
+ transition{j}(i,i+1)=eps;
+ transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-2*eps;
+ else
+ if sum(ind_colc)==0
+ transition{j}(i,i)=eps;
+ transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-2*eps;
+ transition{j}(i,i+2)=eps;
+ else
+ transition{j}(i,i)=eps;
+ transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+ transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+ end
+ end
+ else
+ if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+ if sum(ind_colc)==0
+ transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-2*eps;
+ transition{j}(i,i+1)=eps;
+ transition{j}(i,i+2)=eps;
+ else
+ transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+ transition{j}(i,i+1)=eps;
+ transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+ end
+ else
+ if sum(ind_colc)==0
+ transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+ transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+ transition{j}(i,i+2)=eps;
+ else
+ transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec);
+ transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec);
+ transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec);
+ end
+ end
+ end
+
+ end
+ for k=M:sum(C(1,:))
+ transition{k}(i,:)=transition{M}(i,:);
+ end
+end
+end
+
+
+
+
+% for i=[1,3]
+% M=max(C(:,i));
+% for j=1:M
+% ind_col=C(:,i)>j
+% ind_col_prec=C(:,i)>=j
+% num1=sum(ind_col)
+% num2=(sum(ind_col_prec)-sum(ind_col))
+% if sum(ind_col)==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=num2/sum(ind_col_prec)-eps
+% elseif (sum(ind_col_prec)-sum(ind_col))==0
+% transition{j}(i,i)=num1/sum(ind_col_prec)-eps
+% transition{j}(i,i+1)=eps;
+% else
+% transition{j}(i,i)=num1/sum(ind_col_prec);
+% transition{j}(i,i+1)=num2/sum(ind_col_prec);
+% end
+% end
+% for k=M:sum(C(1,:))
+% transition{k}(i,:)=transition{M}(i,:);
+% end
+% end
+% for i=2
+% M=max(C(:,i));
+% for j=1:M
+% ind_col=C(:,i)>j;
+% ind_col_prec=C(:,i)>=j;
+% ind_colb1=C(:,2)>j;
+% ind_colc=C(:,i)==j;
+% if sum(ind_colb1)==0
+% if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-2*eps;
+% else
+% if sum(ind_colc)==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-2*eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+% end
+% end
+% else
+% if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+% if sum(ind_colc)==0
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-2*eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+% end
+% else
+% if sum(ind_colc)==0
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec);
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec);
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec);
+% end
+% end
+% end
+%
+% end
+% for k=M:sum(C(1,:))
+% transition{k}(i,:)=transition{M}(i,:);
+% end
+% end
+% end
+% % B1 B2 C
+% % 0 0 1
+% % 0 1 0
+% % 0 1 1
+% % 1 0 0
+% % 1 0 1
+% % 1 1 0
+% %
+% % si 0
+% % si 0
+% % sinon
+% % si 0
+% % sinon
+% % sinon
+% % si 0
+% % si 0
+% % sinon
+% % sinon
+% % si 0
+% % sinon
+% % %% nombre de transitions
+% % ind_col=C(:,2)==4 %on trouve les indices
+% % C(ind_col,:) %toutes les lignes de la colonne 2 qui est égale à 4
+% % %% vers C
+% % sum(C(ind_col,3)==0) %on trouve le nombre de transition vers C
+%
+% % %% nombre de non-transitions (reste sur B1)
+% % ind_col=C(:,2)>4 (modifie)
+% %sum(ind_col)
+% %sum(a==2)
diff --git a/parmorceauxMax-git/matrice_transition.m b/parmorceauxMax-git/matrice_transition.m
new file mode 100644
index 0000000000000000000000000000000000000000..3d3a98165a14641bb8ca5437f47dd1935dee5716
--- /dev/null
+++ b/parmorceauxMax-git/matrice_transition.m
@@ -0,0 +1,195 @@
+function transition=matrice_transition(eps)
+%C=xlsread('p_t_test.xls');
+C=xlsread('proba_transition.xls');
+%transition=repmat({[zeros(3,4);[0 0 0 1]]},1,sum(C(1,:)));
+N=sum(C(1,:));
+A=repmat({zeros(1,N)}, 3, 4);
+B=repmat({zeros(1,N)}, 1, 3);
+transition=[A; [B, {ones(1,N)}]];
+for i=[1,3]
+ M=max(C(:,i));
+ for j=1:M
+ ind_col=C(:,i)>j
+ ind_col_prec=C(:,i)>=j
+ num1=sum(ind_col)
+ num2=(sum(ind_col_prec)-sum(ind_col))
+ if sum(ind_col)==0
+ transition{i,i}(j)=eps;
+ transition{i,i+1}(j)=num2/sum(ind_col_prec)-eps
+ elseif (sum(ind_col_prec)-sum(ind_col))==0
+ transition{i,i}(j)=num1/sum(ind_col_prec)-eps
+ transition{i,i+1}(j)=eps;
+ else
+ transition{i,i}(j)=num1/sum(ind_col_prec);
+ transition{i,i+1}(j)=num2/sum(ind_col_prec);
+ end
+ end
+ for k=M:sum(C(1,:))
+ transition{i,i}(k)=transition{i,i}(M);
+ transition{i,i+1}(k)=transition{i,i+1}(M);
+ end
+end
+for i=2
+ M=max(C(:,i));
+ for j=1:M
+ ind_col=C(:,i)>j;
+ ind_col_prec=C(:,i)>=j;
+ ind_colb1=C(:,2)>j;
+ ind_colc=C(:,i)==j;
+ if sum(ind_colb1)==0
+ if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+ transition{i,i}(j)=eps;
+ transition{i,i+1}(j)=eps;
+ transition{i,i+2}(j)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-2*eps;
+ else
+ if sum(ind_colc)==0
+ transition{i,i}(j)=eps;
+ transition{i,i+1}(j)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-2*eps;
+ transition{i,i+2}(j)=eps;
+ else
+ transition{i,i}(j)=eps;
+ transition{i,i+1}(j)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+ transition{i,i+2}(j)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+ end
+ end
+ else
+ if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+ if sum(ind_colc)==0
+ transition{i,i}(j)=sum(ind_colb1)/sum(ind_col_prec)-2*eps;
+ transition{i,i+1}(j)=eps;
+ transition{i,i+2}(j)=eps;
+ else
+ transition{i,i}(j)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+ transition{i,i+1}(j)=eps;
+ transition{i,i+2}(j)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+ end
+ else
+ if sum(ind_colc)==0
+ transition{i,i}(j)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+ transition{i,i+1}(j)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+ transition{i,i+2}(j)=eps;
+ else
+ transition{i,i}(j)=sum(ind_colb1)/sum(ind_col_prec);
+ transition{i,i+2}(j)=sum(C(ind_colc,3)==0)/sum(ind_col_prec);
+ transition{i,i+1}(j)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec);
+ end
+ end
+ end
+
+ end
+ for k=M:sum(C(1,:))
+ transition{i,i}(k)=transition{i,i}(M);
+ transition{i,i+1}(k)=transition{i,i+1}(M);
+ transition{i,i+2}(k)=transition{i,i+2}(M);
+ end
+end
+end
+
+
+
+
+% for i=[1,3]
+% M=max(C(:,i));
+% for j=1:M
+% ind_col=C(:,i)>j
+% ind_col_prec=C(:,i)>=j
+% num1=sum(ind_col)
+% num2=(sum(ind_col_prec)-sum(ind_col))
+% if sum(ind_col)==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=num2/sum(ind_col_prec)-eps
+% elseif (sum(ind_col_prec)-sum(ind_col))==0
+% transition{j}(i,i)=num1/sum(ind_col_prec)-eps
+% transition{j}(i,i+1)=eps;
+% else
+% transition{j}(i,i)=num1/sum(ind_col_prec);
+% transition{j}(i,i+1)=num2/sum(ind_col_prec);
+% end
+% end
+% for k=M:sum(C(1,:))
+% transition{k}(i,:)=transition{M}(i,:);
+% end
+% end
+% for i=2
+% M=max(C(:,i));
+% for j=1:M
+% ind_col=C(:,i)>j;
+% ind_col_prec=C(:,i)>=j;
+% ind_colb1=C(:,2)>j;
+% ind_colc=C(:,i)==j;
+% if sum(ind_colb1)==0
+% if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-2*eps;
+% else
+% if sum(ind_colc)==0
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-2*eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+% end
+% end
+% else
+% if (sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))==0
+% if sum(ind_colc)==0
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-2*eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+1)=eps;
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec)-1/2*eps;
+% end
+% else
+% if sum(ind_colc)==0
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec)-1/2*eps;
+% transition{j}(i,i+2)=eps;
+% else
+% transition{j}(i,i)=sum(ind_colb1)/sum(ind_col_prec);
+% transition{j}(i,i+2)=sum(C(ind_colc,3)==0)/sum(ind_col_prec);
+% transition{j}(i,i+1)=(sum(ind_col_prec)-sum(ind_colb1)-sum(C(ind_colc,3)==0))/sum(ind_col_prec);
+% end
+% end
+% end
+%
+% end
+% for k=M:sum(C(1,:))
+% transition{k}(i,:)=transition{M}(i,:);
+% end
+% end
+% end
+% % B1 B2 C
+% % 0 0 1
+% % 0 1 0
+% % 0 1 1
+% % 1 0 0
+% % 1 0 1
+% % 1 1 0
+% %
+% % si 0
+% % si 0
+% % sinon
+% % si 0
+% % sinon
+% % sinon
+% % si 0
+% % si 0
+% % sinon
+% % sinon
+% % si 0
+% % sinon
+% % %% nombre de transitions
+% % ind_col=C(:,2)==4 %on trouve les indices
+% % C(ind_col,:) %toutes les lignes de la colonne 2 qui est égale à 4
+% % %% vers C
+% % sum(C(ind_col,3)==0) %on trouve le nombre de transition vers C
+%
+% % %% nombre de non-transitions (reste sur B1)
+% % ind_col=C(:,2)>4 (modifie)
+% %sum(ind_col)
+% %sum(a==2)
diff --git a/parmorceauxMax-git/p_t_test.xls b/parmorceauxMax-git/p_t_test.xls
new file mode 100644
index 0000000000000000000000000000000000000000..81c7064328c65c5458228cf34d5b94104be8ed35
Binary files /dev/null and b/parmorceauxMax-git/p_t_test.xls differ
diff --git a/parmorceauxMax-git/prepare_models_part2.m b/parmorceauxMax-git/prepare_models_part2.m
index 0bbd199ec8c8f9ca87317ff69c86f1e863bb4a91..0da3e085bc5f78508f06d53feb6dffa1a2db9b03 100644
--- a/parmorceauxMax-git/prepare_models_part2.m
+++ b/parmorceauxMax-git/prepare_models_part2.m
@@ -1,7 +1,7 @@
% prepare the pdf interpolants for the dynamics model
function prepare_models_part2
-load 'probaC.mat';
+load 'probaA.mat';
num_dim = length(size(Proba_XY));
dims = size(Proba_XY);
@@ -107,6 +107,6 @@ end
interpolant_Y = griddedInterpolant(args, Proba_Y);
%% save
-save('interpolants_C.mat','interpolant_XY','interpolant_Y','P_Yi_knowing_X');
+save('interpolants_A.mat','interpolant_XY','interpolant_Y','P_Yi_knowing_X');
end
diff --git a/parmorceauxMax-git/probaA.mat b/parmorceauxMax-git/probaA.mat
index 8f797162fa60ba960343df01d597402c0b800719..8c8d85ccb89f8f4cedc38b14d9f7c2fab047ad8b 100644
Binary files a/parmorceauxMax-git/probaA.mat and b/parmorceauxMax-git/probaA.mat differ
diff --git a/parmorceauxMax-git/probaB1.mat b/parmorceauxMax-git/probaB1.mat
index e0e76226cb48604889353a8f1c13ad3183785f7d..3e34bce2b0d78f29a4b8a1648362bcd9d386f324 100644
Binary files a/parmorceauxMax-git/probaB1.mat and b/parmorceauxMax-git/probaB1.mat differ
diff --git a/parmorceauxMax-git/probaB2.mat b/parmorceauxMax-git/probaB2.mat
index 519e62351c9472e16d7e70b8015106545eaf400f..9087788a4eb5f20aa719d0415f2698ffee769552 100644
Binary files a/parmorceauxMax-git/probaB2.mat and b/parmorceauxMax-git/probaB2.mat differ
diff --git a/parmorceauxMax-git/probaC.mat b/parmorceauxMax-git/probaC.mat
index 7bc909534af3c86506f791f3c28eea8118274edd..bd6e1383c851afabaa2794b957ce0ac6ea8ec8a1 100644
Binary files a/parmorceauxMax-git/probaC.mat and b/parmorceauxMax-git/probaC.mat differ
diff --git a/parmorceauxMax-git/proba_transition.ods b/parmorceauxMax-git/proba_transition.ods
new file mode 100644
index 0000000000000000000000000000000000000000..dca9a777819f28bcc21718d5e396511c0bf492af
Binary files /dev/null and b/parmorceauxMax-git/proba_transition.ods differ
diff --git a/parmorceauxMax-git/proba_transition.xls b/parmorceauxMax-git/proba_transition.xls
new file mode 100644
index 0000000000000000000000000000000000000000..2e0f00b0118dd7583efdfead67c9fa7ec9b55221
Binary files /dev/null and b/parmorceauxMax-git/proba_transition.xls differ
diff --git a/parmorceauxMax-git/trans.mat b/parmorceauxMax-git/trans.mat
new file mode 100644
index 0000000000000000000000000000000000000000..170b68385ea847fcab2ce7983d9a6f23e510deb5
Binary files /dev/null and b/parmorceauxMax-git/trans.mat differ
diff --git a/training/A/probaAorigine.mat b/training/A/probaAorigine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..8f797162fa60ba960343df01d597402c0b800719
Binary files /dev/null and b/training/A/probaAorigine.mat differ
diff --git a/training/B1/probaB1origine.mat b/training/B1/probaB1origine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..e0e76226cb48604889353a8f1c13ad3183785f7d
Binary files /dev/null and b/training/B1/probaB1origine.mat differ
diff --git a/training/B2/probaB2origine.mat b/training/B2/probaB2origine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..519e62351c9472e16d7e70b8015106545eaf400f
Binary files /dev/null and b/training/B2/probaB2origine.mat differ
diff --git a/training/B2/training/.~lock.points_courbes.xls# b/training/B2/training/.~lock.points_courbes.xls#
new file mode 100644
index 0000000000000000000000000000000000000000..644d1bd25640dafc1f1c77d3d210a84b070c8da9
--- /dev/null
+++ b/training/B2/training/.~lock.points_courbes.xls#
@@ -0,0 +1 @@
+,DESKTOP-IU67DUI/Laurie,DESKTOP-IU67DUI,02.06.2023 17:25,file:///C:/Users/Laurie/AppData/Roaming/LibreOffice/4;
\ No newline at end of file
diff --git a/training/B2/training/A/prepare_models_part1modifie.m b/training/B2/training/A/prepare_models_part1modifie.m
new file mode 100644
index 0000000000000000000000000000000000000000..55c8cfe9edcd75ae14b28335965091097ef70ad7
--- /dev/null
+++ b/training/B2/training/A/prepare_models_part1modifie.m
@@ -0,0 +1,59 @@
+% build dynamics model (part 1)
+%modele A
+Y=load('../ListeDesVolumes.txt');
+y_training=zeros(232,1);%zeros(210,1);%zeros(192,1)
+C=xlsread('../points_courbes.xls');
+
+% on normalise les patients
+J=1;
+I=1;
+ymax_training=Y;
+ymax_training=zeros(22,1);
+while (I<=22);
+ % on cherche le max du patient
+ J=1;
+ while (J<=25);
+ if Y(J,I) > ymax_training(I,1)
+ ymax_training(I,1)=Y(J,I);
+ end
+ J=J+1;
+ end
+
+ % on normalise le patient
+ J=1;
+ while (J<=25);
+ Y(J,I) = Y(J,I) / ymax_training(I,1);
+ J=J+1;
+ end
+
+ I=I+1;
+end
+
+% on récupère les frames qui nous intéressent
+J=1;
+I=1;
+i=1;
+while (I<=22);
+ J=1;
+ while (J<C(I,1));
+ y_training(i,1)=Y(J,I);
+ i=i+1;
+ J=J+1;
+ end
+ I=I+1;
+end
+Y_training = y_training;
+
+%creation de X_training
+k=1;
+X_training=[];
+while k<=22
+ X=[0:1/(C(k,1)-2):1];
+ X_training = [X_training, X];
+ k=k+1;
+end
+X_training=X_training';
+%X_training=repmat(X_training,18,1);
+
+
+save('trainingA.mat','X_training','Y_training')
\ No newline at end of file
diff --git a/training/B2/training/A/probaA.mat b/training/B2/training/A/probaA.mat
new file mode 100644
index 0000000000000000000000000000000000000000..352e728ccb9c5c1b14518d6dace71d0fbd34dad7
Binary files /dev/null and b/training/B2/training/A/probaA.mat differ
diff --git a/training/B2/training/A/probaA_0.06.mat b/training/B2/training/A/probaA_0.06.mat
new file mode 100644
index 0000000000000000000000000000000000000000..8c8d85ccb89f8f4cedc38b14d9f7c2fab047ad8b
Binary files /dev/null and b/training/B2/training/A/probaA_0.06.mat differ
diff --git a/training/B2/training/A/probaAorigine.mat b/training/B2/training/A/probaAorigine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..8f797162fa60ba960343df01d597402c0b800719
Binary files /dev/null and b/training/B2/training/A/probaAorigine.mat differ
diff --git a/training/B2/training/A/trainingA.mat b/training/B2/training/A/trainingA.mat
new file mode 100644
index 0000000000000000000000000000000000000000..02bcb3b79de4777860b4a01f4ac21d33428a9c3b
Binary files /dev/null and b/training/B2/training/A/trainingA.mat differ
diff --git a/training/B2/training/B1/prepare_models_part1modifie.m b/training/B2/training/B1/prepare_models_part1modifie.m
new file mode 100644
index 0000000000000000000000000000000000000000..1188674ab96cad078321d9039ee7c8b99ec21ca9
--- /dev/null
+++ b/training/B2/training/B1/prepare_models_part1modifie.m
@@ -0,0 +1,72 @@
+% build dynamics model (part 1)
+%modele B1
+Y=load('../ListeDesVolumes.txt');
+y_training=zeros(159,1)%zeros(145,1);%zeros(127,1)
+C=xlsread('../points_courbes.xls');
+
+% on normalise les patients
+J=1;
+I=1;
+ymax_training=Y;
+ymax_training=zeros(22,1);
+while (I<=22);
+ % on cherche le max du patient
+ J=1;
+ while (J<=25);
+ if Y(J,I) > ymax_training(I,1)
+ ymax_training(I,1)=Y(J,I);
+ end
+ J=J+1;
+ end
+
+ % on normalise le patient
+ J=1;
+ while (J<=25);
+ Y(J,I) = Y(J,I) / ymax_training(I,1);
+ J=J+1;
+ end
+
+ I=I+1;
+end
+
+% on récupère les frames qui nous intéressent
+J=1;
+I=1;
+i=1;
+while (I<=22); %on parcourt tout les patients
+ J=C(I,1); %on prend que la premiere colonne du .xls
+ if C(I,3)>0 %si le patient a un plateau
+ while (J<C(I,3)); %on parcourt tout les éléments de la 3ieme colonnes J in 1:element troisieme col
+ y_training(i,1)=Y(J,I); %exp patient 1: y_training=une col
+ i=i+1;
+ J=J+1;
+ end
+ else
+ while (J<C(I,2));
+ y_training(i,1)=Y(J,I);
+ i=i+1;
+ J=J+1;
+ end
+ end
+ I=I+1;
+end
+Y_training = y_training;
+
+%creation de X_training
+k=1;
+X_training=[];
+while k<=22
+ if C(k,3)>0
+ X=[0:1/(C(k,3)-C(k,1)-1):1];
+ X_training = [X_training, X];
+ else
+ X=[0:1/(C(k,2)-C(k,1)-1):1];
+ X_training = [X_training, X];
+ end
+ k=k+1;
+end
+X_training=X_training';
+%X_training=repmat(X_training,18,1);
+
+
+save('trainingB1.mat','X_training','Y_training')
\ No newline at end of file
diff --git a/training/B2/training/B1/probaB1.mat b/training/B2/training/B1/probaB1.mat
new file mode 100644
index 0000000000000000000000000000000000000000..3e34bce2b0d78f29a4b8a1648362bcd9d386f324
Binary files /dev/null and b/training/B2/training/B1/probaB1.mat differ
diff --git a/training/B2/training/B1/probaB1origine.mat b/training/B2/training/B1/probaB1origine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..e0e76226cb48604889353a8f1c13ad3183785f7d
Binary files /dev/null and b/training/B2/training/B1/probaB1origine.mat differ
diff --git a/training/B2/training/B1/trainingB1.mat b/training/B2/training/B1/trainingB1.mat
new file mode 100644
index 0000000000000000000000000000000000000000..8408bbebc981a976b4dda294fbb24cb607f67afc
Binary files /dev/null and b/training/B2/training/B1/trainingB1.mat differ
diff --git a/training/B2/training/B2/prepare_models_part1modifie.m b/training/B2/training/B2/prepare_models_part1modifie.m
new file mode 100644
index 0000000000000000000000000000000000000000..3596ecb553b44be7a97f9a893b4a52b36c0c5997
--- /dev/null
+++ b/training/B2/training/B2/prepare_models_part1modifie.m
@@ -0,0 +1,63 @@
+% build dynamics model (part 1)
+%modele B2 %le vrai
+Y=load('../ListeDesVolumes.txt');
+y_training=zeros(88,1);%zeros(88,1);zeros(73,1)
+C=xlsread('../points_courbes.xls');
+
+% on normalise les patients
+J=1;
+I=1;
+ymax_training=Y;
+ymax_training=zeros(22,1);
+while (I<=22);
+ % on cherche le max du patient
+ J=1;
+ while (J<=25);
+ if Y(J,I) > ymax_training(I,1)
+ ymax_training(I,1)=Y(J,I);
+ end
+ J=J+1;
+ end
+
+ % on normalise le patient
+ J=1;
+ while (J<=25);
+ Y(J,I) = Y(J,I) / ymax_training(I,1);
+ J=J+1;
+ end
+
+ I=I+1;
+end
+
+% on récupère les frames qui nous intéressent
+J=1;
+I=1;
+i=1;
+while (I<=22);
+ J=C(I,3);
+ if C(I,3)>0
+ while (J<C(I,2));
+ y_training(i,1)=Y(J,I);
+ i=i+1;
+ J=J+1;
+ end
+ end
+ I=I+1;
+end
+Y_training = y_training;
+
+%creation de X_training
+k=1;
+X_training=[];
+while k<=22
+ if C(k,3)>0
+ X=[0:1/(C(k,2)-C(k,3)-1):1];
+ X_training = [X_training, X];
+ end
+ k=k+1;
+end
+X_training=X_training';
+%X_training=repmat(X_training,18,1);
+
+
+save('trainingB2.mat','X_training','Y_training')
diff --git a/training/B2/training/B2/probaB2.mat b/training/B2/training/B2/probaB2.mat
new file mode 100644
index 0000000000000000000000000000000000000000..9087788a4eb5f20aa719d0415f2698ffee769552
Binary files /dev/null and b/training/B2/training/B2/probaB2.mat differ
diff --git a/training/B2/training/B2/probaB2origine.mat b/training/B2/training/B2/probaB2origine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..519e62351c9472e16d7e70b8015106545eaf400f
Binary files /dev/null and b/training/B2/training/B2/probaB2origine.mat differ
diff --git a/training/B2/training/B2/trainingB2.mat b/training/B2/training/B2/trainingB2.mat
new file mode 100644
index 0000000000000000000000000000000000000000..a0f53a85617114eb1db8ccc145c6cb06c7113df5
Binary files /dev/null and b/training/B2/training/B2/trainingB2.mat differ
diff --git a/training/B2/training/C/prepare_models_part1modifie.m b/training/B2/training/C/prepare_models_part1modifie.m
new file mode 100644
index 0000000000000000000000000000000000000000..a78d790c350e5f9e5c61a637242a9e3f6e0d6603
--- /dev/null
+++ b/training/B2/training/C/prepare_models_part1modifie.m
@@ -0,0 +1,59 @@
+% build dynamics model (part 1)
+%modele C
+Y=load('../ListeDesVolumes.txt');
+y_training=zeros(71,1);%zeros(58,1);
+C=xlsread('../points_courbes.xls');
+
+% on normalise les patients
+J=1;
+I=1;
+ymax_training=Y;
+ymax_training=zeros(22,1);
+while (I<=22);
+ % on cherche le max du patient
+ J=1;
+ while (J<=25);
+ if Y(J,I) > ymax_training(I,1)
+ ymax_training(I,1)=Y(J,I);
+ end
+ J=J+1;
+ end
+
+ % on normalise le patient
+ J=1;
+ while (J<=25);
+ Y(J,I) = Y(J,I) / ymax_training(I,1);
+ J=J+1;
+ end
+
+ I=I+1;
+end
+
+% on récupère les frames qui nous intéressent
+J=1;
+I=1;
+i=1;
+while (I<=22);
+ J=C(I,2);
+ while (J<=25);
+ y_training(i,1)=Y(J,I);
+ i=i+1;
+ J=J+1;
+ end
+ I=I+1;
+end
+Y_training = y_training;
+
+%creation de X_training
+k=1;
+X_training=[];
+while k<=22
+ X=[0:1/(25-C(k,2)):1];
+ X_training = [X_training, X];
+ k=k+1;
+end
+X_training=X_training';
+%X_training=repmat(X_training,18,1);
+
+
+save('trainingC.mat','X_training','Y_training')
\ No newline at end of file
diff --git a/training/B2/training/C/probaC.mat b/training/B2/training/C/probaC.mat
new file mode 100644
index 0000000000000000000000000000000000000000..bd6e1383c851afabaa2794b957ce0ac6ea8ec8a1
Binary files /dev/null and b/training/B2/training/C/probaC.mat differ
diff --git a/training/B2/training/C/probaCorigine.mat b/training/B2/training/C/probaCorigine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..7bc909534af3c86506f791f3c28eea8118274edd
Binary files /dev/null and b/training/B2/training/C/probaCorigine.mat differ
diff --git a/training/B2/training/C/trainingC.mat b/training/B2/training/C/trainingC.mat
new file mode 100644
index 0000000000000000000000000000000000000000..3b9a1037ef2bf355fccd71d4d7a7836f0cea44ae
Binary files /dev/null and b/training/B2/training/C/trainingC.mat differ
diff --git a/training/B2/training/ListeDesVolumes.txt b/training/B2/training/ListeDesVolumes.txt
new file mode 100644
index 0000000000000000000000000000000000000000..b406c7d918304bbbe4e83ef50920ab8c3241dddc
--- /dev/null
+++ b/training/B2/training/ListeDesVolumes.txt
@@ -0,0 +1,26 @@
+2.017900000000000000e+04 4.595000000000000000e+04 2.857200000000000000e+04 3.557800000000000000e+04 2.554000000000000000e+04 3.495200000000000000e+04 4.747500000000000000e+04 4.773200000000000000e+04 3.376800000000000000e+04 2.268700000000000000e+04 1.736900000000000000e+04 2.652400000000000000e+04 2.660500000000000000e+04 2.709200000000000000e+04 3.605400000000000000e+04 3.408500000000000000e+04 4.594800000000000000e+04 2.803600000000000000e+04 1.926700000000000000e+04 2.163700000000000000e+04 1.954400000000000000e+04 2.706000000000000000e+04
+1.976600000000000000e+04 4.366700000000000000e+04 2.741300000000000000e+04 3.383600000000000000e+04 2.484400000000000000e+04 3.148600000000000000e+04 4.420200000000000000e+04 4.456400000000000000e+04 3.171500000000000000e+04 2.247200000000000000e+04 1.652300000000000000e+04 2.640600000000000000e+04 2.531800000000000000e+04 2.706800000000000000e+04 3.484400000000000000e+04 3.201200000000000000e+04 4.275500000000000000e+04 2.691200000000000000e+04 1.902000000000000000e+04 1.804500000000000000e+04 1.934700000000000000e+04 2.273200000000000000e+04
+1.977000000000000000e+04 4.311500000000000000e+04 2.722200000000000000e+04 3.209400000000000000e+04 2.363100000000000000e+04 3.100600000000000000e+04 4.366500000000000000e+04 4.379100000000000000e+04 3.075100000000000000e+04 2.212400000000000000e+04 1.634900000000000000e+04 2.540400000000000000e+04 2.480100000000000000e+04 2.519600000000000000e+04 3.421400000000000000e+04 3.125700000000000000e+04 4.029700000000000000e+04 2.683100000000000000e+04 1.874400000000000000e+04 1.839600000000000000e+04 1.794400000000000000e+04 2.073500000000000000e+04
+1.913400000000000000e+04 4.153000000000000000e+04 2.608300000000000000e+04 3.097800000000000000e+04 2.222000000000000000e+04 3.025500000000000000e+04 4.231700000000000000e+04 4.038100000000000000e+04 2.923700000000000000e+04 2.092100000000000000e+04 1.580100000000000000e+04 2.391200000000000000e+04 2.295800000000000000e+04 2.370600000000000000e+04 3.185700000000000000e+04 2.990300000000000000e+04 3.518600000000000000e+04 2.627300000000000000e+04 1.815700000000000000e+04 1.878900000000000000e+04 1.577000000000000000e+04 1.713100000000000000e+04
+1.780200000000000000e+04 3.887800000000000000e+04 2.408000000000000000e+04 2.880700000000000000e+04 2.035200000000000000e+04 2.911200000000000000e+04 3.969300000000000000e+04 3.804300000000000000e+04 2.593200000000000000e+04 1.870700000000000000e+04 1.485800000000000000e+04 2.147700000000000000e+04 2.064300000000000000e+04 2.103900000000000000e+04 2.948100000000000000e+04 2.747100000000000000e+04 3.160000000000000000e+04 2.509800000000000000e+04 1.702000000000000000e+04 1.517800000000000000e+04 1.431800000000000000e+04 1.408500000000000000e+04
+1.588600000000000000e+04 3.541100000000000000e+04 2.122600000000000000e+04 2.446700000000000000e+04 1.797300000000000000e+04 2.772700000000000000e+04 3.738800000000000000e+04 3.593300000000000000e+04 2.261800000000000000e+04 1.602300000000000000e+04 1.337000000000000000e+04 1.926400000000000000e+04 1.665200000000000000e+04 1.791700000000000000e+04 2.687900000000000000e+04 2.287100000000000000e+04 2.690200000000000000e+04 2.303000000000000000e+04 1.368000000000000000e+04 1.375100000000000000e+04 1.279900000000000000e+04 1.138800000000000000e+04
+1.366500000000000000e+04 3.127400000000000000e+04 1.778300000000000000e+04 2.124500000000000000e+04 1.536700000000000000e+04 2.518900000000000000e+04 3.514200000000000000e+04 3.325000000000000000e+04 1.964100000000000000e+04 1.269400000000000000e+04 1.142100000000000000e+04 1.642100000000000000e+04 1.365800000000000000e+04 1.312300000000000000e+04 2.359400000000000000e+04 1.871000000000000000e+04 2.273600000000000000e+04 2.048200000000000000e+04 1.148600000000000000e+04 1.200000000000000000e+04 1.091300000000000000e+04 9.589000000000000000e+03
+1.152300000000000000e+04 2.777600000000000000e+04 1.440000000000000000e+04 1.777900000000000000e+04 1.330700000000000000e+04 2.091400000000000000e+04 3.266200000000000000e+04 2.881200000000000000e+04 1.677100000000000000e+04 1.034000000000000000e+04 9.695000000000000000e+03 1.379600000000000000e+04 1.103800000000000000e+04 9.875000000000000000e+03 2.001800000000000000e+04 1.541500000000000000e+04 2.007000000000000000e+04 1.818400000000000000e+04 9.703000000000000000e+03 1.040900000000000000e+04 9.179000000000000000e+03 8.362000000000000000e+03
+1.003000000000000000e+04 2.468400000000000000e+04 1.178500000000000000e+04 1.452600000000000000e+04 1.196400000000000000e+04 1.764800000000000000e+04 2.837900000000000000e+04 2.282100000000000000e+04 1.454600000000000000e+04 8.903000000000000000e+03 7.768000000000000000e+03 1.226800000000000000e+04 9.453000000000000000e+03 8.311000000000000000e+03 1.770300000000000000e+04 1.241400000000000000e+04 1.909400000000000000e+04 1.537200000000000000e+04 8.250000000000000000e+03 9.332000000000000000e+03 7.974000000000000000e+03 7.824000000000000000e+03
+8.939000000000000000e+03 2.226200000000000000e+04 1.016900000000000000e+04 1.136000000000000000e+04 1.133500000000000000e+04 1.578100000000000000e+04 2.429100000000000000e+04 2.012000000000000000e+04 1.332800000000000000e+04 7.947000000000000000e+03 6.748000000000000000e+03 1.127000000000000000e+04 8.820000000000000000e+03 8.105000000000000000e+03 1.631400000000000000e+04 1.118600000000000000e+04 1.637800000000000000e+04 1.354900000000000000e+04 7.261000000000000000e+03 9.682000000000000000e+03 7.012000000000000000e+03 8.599000000000000000e+03
+8.192000000000000000e+03 2.044700000000000000e+04 9.667000000000000000e+03 1.076000000000000000e+04 1.131500000000000000e+04 1.463400000000000000e+04 2.121800000000000000e+04 1.976800000000000000e+04 1.303400000000000000e+04 7.424000000000000000e+03 6.289000000000000000e+03 1.079500000000000000e+04 8.845000000000000000e+03 8.931000000000000000e+03 1.553600000000000000e+04 1.103600000000000000e+04 1.719600000000000000e+04 1.234300000000000000e+04 6.785000000000000000e+03 9.296000000000000000e+03 6.230000000000000000e+03 7.907000000000000000e+03
+7.745000000000000000e+03 1.927200000000000000e+04 9.852000000000000000e+03 1.119700000000000000e+04 1.201900000000000000e+04 1.407400000000000000e+04 1.986100000000000000e+04 2.125000000000000000e+04 1.384100000000000000e+04 7.323000000000000000e+03 6.200000000000000000e+03 1.068400000000000000e+04 9.161000000000000000e+03 1.288900000000000000e+04 1.533000000000000000e+04 1.201600000000000000e+04 1.959900000000000000e+04 1.138500000000000000e+04 6.931000000000000000e+03 9.940000000000000000e+03 5.746000000000000000e+03 8.832000000000000000e+03
+7.575000000000000000e+03 1.943300000000000000e+04 1.038700000000000000e+04 1.426500000000000000e+04 1.351200000000000000e+04 1.436000000000000000e+04 1.945500000000000000e+04 2.523000000000000000e+04 1.579800000000000000e+04 7.685000000000000000e+03 6.378000000000000000e+03 1.111100000000000000e+04 1.021800000000000000e+04 1.801300000000000000e+04 1.595300000000000000e+04 1.533400000000000000e+04 2.504400000000000000e+04 1.035600000000000000e+04 7.017000000000000000e+03 1.015600000000000000e+04 5.827000000000000000e+03 9.581000000000000000e+03
+8.001000000000000000e+03 1.979900000000000000e+04 1.164100000000000000e+04 1.824900000000000000e+04 1.578100000000000000e+04 1.531400000000000000e+04 2.024800000000000000e+04 3.277700000000000000e+04 1.962600000000000000e+04 9.002000000000000000e+03 6.719000000000000000e+03 1.265200000000000000e+04 1.425800000000000000e+04 2.191300000000000000e+04 1.729700000000000000e+04 2.024100000000000000e+04 3.037400000000000000e+04 1.000900000000000000e+04 7.391000000000000000e+03 1.024000000000000000e+04 6.672000000000000000e+03 1.205800000000000000e+04
+9.417000000000000000e+03 2.048000000000000000e+04 1.421300000000000000e+04 2.230700000000000000e+04 1.858300000000000000e+04 1.749200000000000000e+04 2.238800000000000000e+04 3.625900000000000000e+04 2.428500000000000000e+04 1.196400000000000000e+04 7.334000000000000000e+03 1.498500000000000000e+04 1.737500000000000000e+04 2.384200000000000000e+04 2.052400000000000000e+04 2.562000000000000000e+04 3.373900000000000000e+04 1.007100000000000000e+04 8.095000000000000000e+03 1.074500000000000000e+04 7.861000000000000000e+03 1.363100000000000000e+04
+1.211100000000000000e+04 2.146100000000000000e+04 1.697100000000000000e+04 2.583100000000000000e+04 2.038400000000000000e+04 2.173900000000000000e+04 2.602000000000000000e+04 3.852800000000000000e+04 2.717400000000000000e+04 1.442900000000000000e+04 8.389000000000000000e+03 1.744200000000000000e+04 1.917800000000000000e+04 2.479900000000000000e+04 2.386800000000000000e+04 2.824300000000000000e+04 3.646800000000000000e+04 1.071700000000000000e+04 9.878000000000000000e+03 1.157100000000000000e+04 1.021100000000000000e+04 1.511600000000000000e+04
+1.446000000000000000e+04 2.429300000000000000e+04 1.935800000000000000e+04 2.742300000000000000e+04 2.158600000000000000e+04 2.448200000000000000e+04 3.097000000000000000e+04 4.051400000000000000e+04 2.902500000000000000e+04 1.618700000000000000e+04 1.061000000000000000e+04 2.043600000000000000e+04 2.030300000000000000e+04 2.516600000000000000e+04 2.609400000000000000e+04 2.945100000000000000e+04 3.795000000000000000e+04 1.262400000000000000e+04 1.243700000000000000e+04 1.227300000000000000e+04 1.172300000000000000e+04 1.646000000000000000e+04
+1.609800000000000000e+04 2.818800000000000000e+04 2.095900000000000000e+04 2.868300000000000000e+04 2.231800000000000000e+04 2.652500000000000000e+04 3.429200000000000000e+04 4.378500000000000000e+04 3.026900000000000000e+04 1.727900000000000000e+04 1.222300000000000000e+04 2.228500000000000000e+04 2.084300000000000000e+04 2.544700000000000000e+04 2.758700000000000000e+04 3.022500000000000000e+04 3.880600000000000000e+04 1.486100000000000000e+04 1.421700000000000000e+04 1.375700000000000000e+04 1.356200000000000000e+04 1.784500000000000000e+04
+1.707000000000000000e+04 3.126500000000000000e+04 2.213600000000000000e+04 2.933400000000000000e+04 2.268500000000000000e+04 2.787100000000000000e+04 3.586500000000000000e+04 4.428500000000000000e+04 3.109500000000000000e+04 1.769900000000000000e+04 1.398700000000000000e+04 2.340300000000000000e+04 2.115400000000000000e+04 2.567700000000000000e+04 2.868600000000000000e+04 3.083100000000000000e+04 3.928900000000000000e+04 1.764100000000000000e+04 1.539900000000000000e+04 1.443800000000000000e+04 1.471400000000000000e+04 1.958700000000000000e+04
+1.750200000000000000e+04 3.430900000000000000e+04 2.331000000000000000e+04 2.987000000000000000e+04 2.290300000000000000e+04 2.880500000000000000e+04 3.845000000000000000e+04 4.465800000000000000e+04 3.127400000000000000e+04 1.776000000000000000e+04 1.486200000000000000e+04 2.410900000000000000e+04 2.122800000000000000e+04 2.582800000000000000e+04 2.951300000000000000e+04 3.111400000000000000e+04 3.991600000000000000e+04 1.927100000000000000e+04 1.631200000000000000e+04 1.507700000000000000e+04 1.574300000000000000e+04 2.142300000000000000e+04
+1.775100000000000000e+04 3.675500000000000000e+04 2.418400000000000000e+04 3.035000000000000000e+04 2.306500000000000000e+04 2.971900000000000000e+04 4.040700000000000000e+04 4.492500000000000000e+04 3.144300000000000000e+04 1.781600000000000000e+04 1.549200000000000000e+04 2.462700000000000000e+04 2.130100000000000000e+04 2.601000000000000000e+04 2.996100000000000000e+04 3.143500000000000000e+04 4.074800000000000000e+04 2.083100000000000000e+04 1.698000000000000000e+04 1.559700000000000000e+04 1.641500000000000000e+04 2.266700000000000000e+04
+1.781000000000000000e+04 3.941300000000000000e+04 2.476700000000000000e+04 3.100800000000000000e+04 2.335100000000000000e+04 3.056900000000000000e+04 4.343400000000000000e+04 4.528800000000000000e+04 3.153200000000000000e+04 1.814000000000000000e+04 1.592200000000000000e+04 2.494900000000000000e+04 2.158400000000000000e+04 2.601200000000000000e+04 3.046100000000000000e+04 3.159700000000000000e+04 4.103900000000000000e+04 2.232500000000000000e+04 1.761900000000000000e+04 1.641400000000000000e+04 1.699600000000000000e+04 2.374000000000000000e+04
+1.801400000000000000e+04 4.223500000000000000e+04 2.544300000000000000e+04 3.223000000000000000e+04 2.370300000000000000e+04 3.164900000000000000e+04 4.511300000000000000e+04 4.545400000000000000e+04 3.203500000000000000e+04 1.884200000000000000e+04 1.640500000000000000e+04 2.532000000000000000e+04 2.229400000000000000e+04 2.610900000000000000e+04 3.130100000000000000e+04 3.210700000000000000e+04 4.169200000000000000e+04 2.393300000000000000e+04 1.825400000000000000e+04 1.733300000000000000e+04 1.748800000000000000e+04 2.445000000000000000e+04
+1.836400000000000000e+04 4.408400000000000000e+04 2.648700000000000000e+04 3.323900000000000000e+04 2.419500000000000000e+04 3.288800000000000000e+04 4.658000000000000000e+04 4.588300000000000000e+04 3.225000000000000000e+04 2.003500000000000000e+04 1.671900000000000000e+04 2.581900000000000000e+04 2.373700000000000000e+04 2.639100000000000000e+04 3.286800000000000000e+04 3.293500000000000000e+04 4.325400000000000000e+04 2.564500000000000000e+04 1.882700000000000000e+04 1.793800000000000000e+04 1.842300000000000000e+04 2.505300000000000000e+04
+1.923300000000000000e+04 4.511000000000000000e+04 2.760900000000000000e+04 3.525700000000000000e+04 2.484700000000000000e+04 3.442000000000000000e+04 4.716400000000000000e+04 4.692700000000000000e+04 3.345200000000000000e+04 2.150200000000000000e+04 1.704100000000000000e+04 2.629300000000000000e+04 2.566600000000000000e+04 2.677800000000000000e+04 3.511800000000000000e+04 3.363500000000000000e+04 4.536500000000000000e+04 2.706900000000000000e+04 1.920400000000000000e+04 1.841100000000000000e+04 1.881800000000000000e+04 2.541000000000000000e+04
+
diff --git a/training/B2/training/points_courbes.xls b/training/B2/training/points_courbes.xls
new file mode 100644
index 0000000000000000000000000000000000000000..6ba22fd27d388651b2ebf22ebd13cebe85332159
Binary files /dev/null and b/training/B2/training/points_courbes.xls differ
diff --git a/training/C/probaCorigine.mat b/training/C/probaCorigine.mat
new file mode 100644
index 0000000000000000000000000000000000000000..7bc909534af3c86506f791f3c28eea8118274edd
Binary files /dev/null and b/training/C/probaCorigine.mat differ