53 lines
1.3 KiB
Mathematica
53 lines
1.3 KiB
Mathematica
|
function self = sample(self,N)
|
||
|
|
%SAMPLE Draw a sample in an optimizer object and instantiates parameter
|
||
|
|
%
|
||
|
|
% Q = SAMPLE(P,N) generates concatenated instantiated optimizer objects
|
||
|
|
% where samples have been drawn for all parameters with associated
|
||
|
|
% samplers.
|
||
|
|
%
|
||
|
|
% INPUT
|
||
|
|
% P : OPTIMIZER object
|
||
|
|
% N : Number of samples to draw
|
||
|
|
%
|
||
|
|
% OUTPUT
|
||
|
|
% Q : OPTIMIZER object
|
||
|
|
%
|
||
|
|
% EXAMPLE
|
||
|
|
% sdpvar x(2,1) w(2,1)
|
||
|
|
% F = [-10 <= x + w <= 10, uncertain(w,'unif',2,3)]
|
||
|
|
% P = optimizer(F,sum(x),[],w,x)
|
||
|
|
% Q = sample(P,10);
|
||
|
|
% plot(Q);
|
||
|
|
% xoptimal = Q([]);
|
||
|
|
%
|
||
|
|
% See also OPTIMIZER, UNCERTAIN
|
||
|
|
|
||
|
|
if nargin < 2
|
||
|
|
N = 1;
|
||
|
|
end
|
||
|
|
allSamples = {};
|
||
|
|
for k = 1:N
|
||
|
|
cells = cell(1,length(self.diminOrig));
|
||
|
|
for i = 1:length(self.diminOrig)
|
||
|
|
if isfield(self.input,'stochastics')
|
||
|
|
if ~isempty(self.input.stochastics{i});
|
||
|
|
temp = {self.input.stochastics{i}.name,self.input.stochastics{i}.parameters{:},self.diminOrig{i}};
|
||
|
|
sampledData = feval(temp{:});
|
||
|
|
cells{i} = sampledData;
|
||
|
|
else
|
||
|
|
cells{i} = [];
|
||
|
|
end
|
||
|
|
end
|
||
|
|
end
|
||
|
|
if ~isempty(cells)
|
||
|
|
y.type = '{}';
|
||
|
|
y.subs = {cells{:},'nosolve'};
|
||
|
|
allSamples{end + 1} = subsref(self,y);
|
||
|
|
else
|
||
|
|
return
|
||
|
|
end
|
||
|
|
end
|
||
|
|
self = horzcat(allSamples{:});
|
||
|
|
|
||
|
|
|