void UnifiedModel::getParameterVectorMask(const std::set<std::string> selected_values_labels, VectorXi& selected_mask) const
{
selected_mask.resize(getParameterVectorAllSize());
selected_mask.fill(0);
int offset = 0;
int size;
// Centers
size = centers_.size()*centers_[0].size();
if (selected_values_labels.find("centers")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(1);
offset += size;
// Widths
size = covars_.size()*covars_[0].cols();
if (selected_values_labels.find("widths")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(2);
offset += size;
// Offsets
size = offsets_.size();
if (selected_values_labels.find("offsets")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(3);
offset += size;
// Slopes
size = slopes_.size()*slopes_[0].size();
if (selected_values_labels.find("slopes")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(4);
offset += size;
assert(offset == getParameterVectorAllSize());
}
void ModelParametersLWPR::getParameterVectorMask(const std::set<std::string> selected_values_labels, VectorXi& selected_mask) const
{
selected_mask.resize(getParameterVectorAllSize());
selected_mask.fill(0);
int offset = 0;
int size;
// Centers
size = n_centers_;
if (selected_values_labels.find("centers")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(1);
offset += size;
// Widths
size = n_widths_;
if (selected_values_labels.find("widths")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(2);
offset += size;
// Offsets
size = n_offsets_;
if (selected_values_labels.find("offsets")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(3);
offset += size;
// Slopes
size = n_slopes_;
if (selected_values_labels.find("slopes")!=selected_values_labels.end())
selected_mask.segment(offset,size).fill(4);
offset += size;
assert(offset == getParameterVectorAllSize());
}
void online_test(const vector<vector<MatrixXd> >& vec_data, const VectorXi& vec_label, const vector<MatrixXd>& weight_0, const MatrixXd& bias_0, const vector<MatrixXd>& weight_1, const MatrixXd& bias_1, const vector<MatrixXd>& weight_2, const MatrixXd& bias_2, const MatrixXd& weight_class, const MatrixXd& bias_class, const int& num_kerns1, const int& num_kerns2, const int& num_kerns3, const int& kern_size, const int& pool_size)
{
vector<vector<MatrixXd> > test_data;
copy(vec_data.begin(), vec_data.begin() + 1, back_inserter(test_data));
VectorXi test_label;
test_label = vec_label.segment(0, 1200);
ConvPoolLayer *p_conv0 = new ConvPoolLayer(test_data, num_kerns1, kern_size, "same", "tanh", weight_0, bias_0);
Pool *p_pool0 = new Pool(p_conv0->batch_maps_activated, pool_size, pool_size);
ConvPoolLayer *p_conv1 = new ConvPoolLayer(p_pool0->output_batch_pooled, num_kerns2, kern_size, "same", "tanh", weight_1, bias_1);
Pool *p_pool1 = new Pool(p_conv1->batch_maps_activated, pool_size, pool_size);
ConvPoolLayer *p_conv2 = new ConvPoolLayer(p_pool1->output_batch_pooled, num_kerns3, kern_size, "same", "tanh", weight_2, bias_2);
Pool *p_pool2 = new Pool(p_conv2->batch_maps_activated, pool_size, pool_size);
MatrixXd feature_vectors;
get_feature_vector(p_pool2->output_batch_pooled, feature_vectors);
// classifier.input = feature_vectors;
Softmax classifier(feature_vectors, 9, test_label, 1, 1, weight_class, bias_class);
classifier.calculation_output();
cout << classifier.m.transpose() << endl;
int accuracy = 0;
for(int i = 0; i < test_label.size(); i++)
{
if(classifier.m(i) == test_label(i))
{
accuracy ++;
}
}
cout << "accuracy : " << accuracy << endl;
}
void QPSolver::convertToConicProblem(ConicSolver& solver)
{
const MatrixXd& F = mLhs;
int newNumVar = mNumVar + 1 + 1 + F.rows(); //x,v,w,t
int newNumCon = mNumCon + F.rows() + 1 + 2;
VectorXd newC = VectorXd::Zero(newNumVar);
newC.head(mNumVar) = mc;
newC(mNumVar) = 1.0;
solver.SetLinearTerm(newC);
MatrixXd newA = MatrixXd::Zero(newNumCon, newNumVar);
VectorXd newLbc = VectorXd::Zero(newNumCon);
VectorXd newUbc = VectorXd::Zero(newNumCon);
VectorXi newIsConstraintLowerBounded = VectorXi::Zero(newNumCon);
VectorXi newIsConstraintUpperBounded = VectorXi::Zero(newNumCon);
if (mNumCon)
{
newA.block(0, 0, mNumCon, mNumVar) = mA;
newLbc.head(mNumCon) = mlbc;
newUbc.head(mNumCon) = mubc;
newIsConstraintLowerBounded.head(mNumCon) = mbConstraintLowerBounded;
newIsConstraintUpperBounded.head(mNumCon) = mbConstraintUpperBounded;
}
newA.block(mNumCon, 0, F.rows(), mNumVar) = F;
newA.block(mNumCon, newNumVar - F.rows(), F.rows(), F.rows()) = -MatrixXd::Identity(F.rows(), F.rows());
newA(mNumCon + F.rows(), mNumVar + 1) = 1.0;
newA(mNumCon + F.rows() + 1, mNumVar) = 1.0;
newA(mNumCon + F.rows() + 2, mNumVar + 1) = 1.0;
newLbc.segment(mNumCon, F.rows()) = VectorXd::Zero(F.rows());
newLbc(mNumCon + F.rows()) = 1.0;
newUbc.segment(mNumCon, F.rows()) = VectorXd::Zero(F.rows());
newUbc(mNumCon + F.rows()) = 1.0;
newIsConstraintLowerBounded.segment(mNumCon, F.rows()) = VectorXi::Constant(F.rows(), 1);
newIsConstraintLowerBounded.tail(3) = VectorXi::Constant(3, 1);
newIsConstraintUpperBounded.segment(mNumCon, F.rows()) = VectorXi::Constant(F.rows(), 1);
newIsConstraintUpperBounded(mNumCon + F.rows()) = 1;
solver.SetConstraints(newA, newLbc, newIsConstraintLowerBounded, newUbc, newIsConstraintUpperBounded);
VectorXd newLb = VectorXd::Zero(newNumVar);
VectorXd newUb = VectorXd::Zero(newNumVar);
VectorXi newBLowerBounded = VectorXi::Zero(newNumVar);
VectorXi newBUpperBounded = VectorXi::Zero(newNumVar);
solver.SetLowerBounds(newLb, newBLowerBounded);
solver.SetUpperBounds(newUb, newBUpperBounded);
Cone cone;
for (int i = mNumVar; i < newNumVar; ++i)
{
cone.mSubscripts.push_back(i);
}
solver.AddCone(cone);
}
