demo1.m

%%
close all; 
addpath(genpath(pwd));
%% Datasets
data_names={'A3','S1','UB','2G','Spiral','Jain','3Circles','AGG','Flame','GaSpCi','GaSpCiNo','1D-EqSp','data_TB_100000','data_SF_100000','data_CC_100000', 'data_CG_100000', 'data_Flower_100000','data_TB_1000000','data_SF_1000000','data_CC_1000000','data_CG_1000000','data_Flower_1000000','One_Dim_uniform_data_1048576','data_TB_10000000'}; 
 

%% Methods
method_names = {'FastLDPMST'};   
%% Start Testing
record_num = 0;
for name_id=1:length(data_names)
    %% load dataset
    clear data annotation_data
    dataName = data_names{name_id};
    disp([num2str(name_id),', ',dataName,':'])
    [data,annotation_data,nC,dataName] = load_data(dataName);
    [N,dim]=size(data);
    
    %% parameter setting
    ratio = 0.01; %  [0.01,0.02] is recommended; not needed for manual cutting; 
    MinSize=ratio*N; % Note: parameter MinSize (i.e.,the minimal cluster size) is dependent on ratio;
    K = ceil(log2(N)); 
     %% compare different methods
    for method_id = 1:length(method_names)
        method = method_names{method_id};
        switch method
            case 'FastLDPMST'  
                [Label,time] = FastLDPMST(data, nC, MinSize, K); %% nC: number of clusters;
            otherwise
                error('method is not included...please name the method appropriately.')
        end
        %% evaluate result and plot
        % diff_colors = linspecer(length(unique(Label))); 
        diff_colors = distinguishable_colors(length(unique(Label)));
        if dim == 2 || dim == 1
            if dim == 1
                data = [data zeros(N,1)];
            end
            figure; idx = 1:N;
            if N>100000,idx = randperm(N,10000);end
            subplot(1,2,1);scatter(data(idx,1),data(idx,2),10,'k','filled'); axis tight; set(gca,'xtick',[],'ytick',[],'FontSize',10,'Linewidth',.01);box on;
            subplot(1,2,2);scatter(data(idx,1),data(idx,2),3,diff_colors(Label(idx),:),'filled'); axis tight; set(gca,'xtick',[],'ytick',[],'FontSize',10,'Linewidth',.01);box on;
			disp('save plot result as a .png file')
			saveas(gcf,['demo1_result_',dataName,'.png'])
        end
        
        record_num = record_num + 1;
        Result_all(record_num).dataName = dataName;
        Result_all(record_num).N = N;
        Result_all(record_num).DIM = dim;
        if exist('annotation_data','var')
            disp('evaluate the clustering result...')
            if any(isnan(annotation_data))
                id_Nan = isnan(annotation_data);
                annotation_data = annotation_data(~id_Nan);
                Label = Label(~id_Nan);
            end
            [NMI,ARI]= NMI_ARI(Label,annotation_data);
            Result_all(record_num).NMI = round(NMI*1000)/1000; 
            Result_all(record_num).ARI = round(ARI*1000)/1000;  
        end
        Result_all(record_num).time = [sprintf('%.3f',time),'sec']; 
        % Result_all(record_num).method = method;
        disp([Result_all(record_num).dataName, ',  Runtime = ',sprintf('%.3f',time),'sec'])
		disp(" "); disp(" ");
    end
 
end
%% show all results
disp(' ******************** All Results ************************ ')
if exist('Result_all','var')
    disp(struct2table(Result_all, 'AsArray', true)) 
end