Dynamic-Calibration/utils/YALMIP-master/extras/savesdpafile.m

125 lines
3.2 KiB
Matlab
Executable File

function solution = savesdpafile(varargin)
%SAVESDPAFILE Saves a problem definition in the SDPA format
%
% SAVESDPAFILE(F,h,'filename') Saves the problem min(h(x)), F(x)>0 to the file filename
% SAVESDPAFILE(F,h) A "Save As" - box will be opened
%
% Note the the SDPA format does not support SOCPs or equalities.
% Equalities will be eliminated using double-sided inequalities. If the
% model contains SOCP constraints the command will exit.
F = varargin{1};
h = varargin{2};
nvars = yalmip('nvars');
if isa(F,'constraint')
F = lmi(F);
end
if any(is(F,'socp'))
error('savesdpafile does not support SOCPs (not supported in the SDPA format');
end
% Expand nonlinear operators
[F,failure,cause] = expandmodel(F,h,sdpsettings);
if failure % Convexity propgation failed
interfacedata = [];
recoverdata = [];
solver = '';
diagnostic.solvertime = 0;
diagnostic.problem = 14;
diagnostic.info = yalmiperror(14,cause);
return
end
% Convert equalities to inequalities
feq =find(is(F,'equality'));
if ~isempty(feq)
f = sdpvar(F(feq));
F(feq)=[];
F = [F, f >= 0, f <= 0];
end
% Get the SP format
[F_struc,K] = lmi2sedumistruct(F);
% Convert the objective
if isempty(h)
c=zeros(nvars,1);
else
[n,m]=size(h);
if ~((n==1) & (m==1))
error('Scalar expression to minimize please.');
else
lmi_variables = getvariables(h);
c = zeros(nvars,1);
for i=1:length(lmi_variables)
c(lmi_variables(i))=getbasematrix(h,lmi_variables(i));
end;
end
end
% Which sdpvar variables are actually in the problem
used_variables_LMI = find(any(F_struc(:,2:end),1));
used_variables_obj = find(any(c',1));
used_variables = uniquestripped([used_variables_LMI used_variables_obj]);
% Check for unbounded variables
unbounded_variables = setdiff(used_variables_obj,used_variables_LMI);
if ~isempty(unbounded_variables)
% Remove unbounded variable from problem
used_variables = setdiff(used_variables,unbounded_variables);
end
% Pick out the necessary rows
if length(used_variables)<nvars
c = c(used_variables);
F_struc = sparse(F_struc(:,[1 1+[used_variables]]));
end
if K.f>0
% Extract the inequalities
A_equ = F_struc(1:K.f,2:end);
b_equ = -F_struc(1:K.f,1);
[Q,R] = qr(A_equ');
n = max(find(any(R')));
Q1 = Q(:,1:n);
Q2 = Q(:,n+1:end);
R = R(1:n,:);
x_equ = Q1*(R'\b_equ);
% Exit if no consistent solution exist
if (norm(A_equ*x_equ-b_equ)>sqrt(eps))
error('Linear constraints inconsistent.');
return
end
% So we dont need these rows anymore
F_struc = F_struc(K.f+1:end,:);
K.f = 0;
% OK, we found a new basis
H = Q2;
% objective in new basis
c = H'*c;
% LMI in new basis
F_struc = [F_struc*[1;x_equ] F_struc(:,2:end)*H];
end
% Is a filename supplied
if nargin<3
[filename, pathname] = uiputfile('*.dat-s', 'Save SDPA sparse format file');
if isa(filename,'double')
return % User canceled
else
% Did the user change the extension
if isempty(findstr(filename,'.'))
filename = [pathname filename '.dat-s'];
else
filename = [pathname filename];
end
end
else
filename = varargin{3};
end
% Save to file
integer_variables = find(ismember(used_variables,yalmip('intvariables')));
createsdplibfile(F_struc, K, c, filename,integer_variables);