function G = gsp_sphere(N, param)
%GSP_SPHERE Create a spherical-shaped graph
% Usage : G = gsp_sphere();
% G = gsp_sphere( param );
%
% Input parameters:
% param : Structure of optional parameters
%
% Output parameters:
% G : Resulting graph
%
% 'gsp_sphere( param )' creates a graph from points sampled on a
% hyper-sphere. The dimension of the sphere can be passed as a parameter.
% It can be sampled in a uniform voxel grid or randomly.
%
% Additional parameters
% ---------------------
%
% param.radius : float the radius of the sphere
% param.nb_pts : int the number of vertices
% param.nb_dim : int the dimension
% param.sampling : ['random'] the variance of the distance kernel
%
% Example:
%
% G = gsp_sphere();
% gsp_plot_graph(G);
% axis square
%
% See also: gsp_cube
%
%
% Url: https://epfl-lts2.github.io/gspbox-html/doc/graphs/gsp_sphere.html
% Copyright (C) 2013-2016 Nathanael Perraudin, Johan Paratte, David I Shuman.
% This file is part of GSPbox version 0.7.5
%
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% 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 General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with this program. If not, see <http://www.gnu.org/licenses/>.
% If you use this toolbox please kindly cite
% N. Perraudin, J. Paratte, D. Shuman, V. Kalofolias, P. Vandergheynst,
% and D. K. Hammond. GSPBOX: A toolbox for signal processing on graphs.
% ArXiv e-prints, Aug. 2014.
% http://arxiv.org/abs/1408.5781
% Author : Johan Paratte
if nargin < 1
N = 300;
end
if nargin < 2
param = {};
end
%Parameters
if ~isfield(param, 'radius'), param.radius = 1; end
if ~isfield(param, 'nb_dim'), param.nb_dim = 3; end
if ~isfield(param, 'sampling'), param.sampling = 'random'; end
K = param.nb_dim;
switch param.sampling
case 'uniform'
% Recursive zonal equal area partition EQ(d, N)
pts = eq_point_set(K - 1, N)';
case 'random'
% Draw angles randomly
% angles = rand(N, K-1);
% if (K > 2)
% angles(:,1:K-2) = angles(:,1:K-2)*2*pi;
% angles(:,K-1) = angles(:,K-1)*pi;
% else
% angles(:,K-1) = angles(:,K-1)*2*pi;
% end
% angles(:,K-1) = angles(:,K-1)*pi;
% pts = ones(N, K);
% % x_1 = r
% pts(:,1) = pts(:,1) .* param.radius;
% for k = 2:K-1
% pts(:,k) = pts(:,k-1) .* sin(angles(:,k-1));
% end
% pts(:,K) = pts(:,K-1);
% pts(:,1:K-1) = pts(:,1:K-1) .* cos(angles);
% pts(:,K) = pts(:,K) .* sin(angles(:,K-1));
pts = normrnd(0, 1, N, K);
for ii = 1:N
pts(ii,:) = pts(ii,:) ./ norm(pts(ii,:));
end
%TODO Implement http://en.wikipedia.org/wiki/Hypersphere#CITEREFMarsaglia1972
otherwise
error('Unknown sampling !');
end
param.type = 'knn';
param.k = 10;
G = gsp_nn_graph(pts, param);
end