void get_samples()
{
ASSERT(samples.size() == fetchers.size());
for (size_t i = 0; i < samples.size(); ++i)
{
CDenseFeatures<float64_t> *ptr = (CDenseFeatures<float64_t>*)fetchers[i]->fetch(samples[i].ptr);
ptr->get_feature_matrix().display_matrix();
}
}
int main()
{
const int32_t feature_cache=0;
const int32_t kernel_cache=0;
const float64_t rbf_width=10;
const float64_t svm_C=10;
const float64_t svm_eps=0.001;
init_shogun();
gen_rand_data();
// create train labels
CLabels* labels=new CLabels(SGVector<float64_t>(lab, NUM));
// create train features
CDenseFeatures<float64_t>* features = new CDenseFeatures<float64_t>(feature_cache);
SG_REF(features);
features->set_feature_matrix(feat);
// create gaussian kernel
CGaussianKernel* kernel = new CGaussianKernel(kernel_cache, rbf_width);
SG_REF(kernel);
kernel->init(features, features);
// create svm via libsvm and train
CLibSVM* svm = new CLibSVM(svm_C, kernel, labels);
SG_REF(svm);
svm->set_epsilon(svm_eps);
svm->train();
printf("num_sv:%d b:%f\n", svm->get_num_support_vectors(), svm->get_bias());
// classify + display output
CLabels* out_labels=svm->apply();
for (int32_t i=0; i<NUM; i++)
printf("out[%d]=%f\n", i, out_labels->get_label(i));
SG_UNREF(out_labels);
SG_UNREF(kernel);
SG_UNREF(features);
SG_UNREF(svm);
exit_shogun();
return 0;
}
SGMatrix<float64_t> CDimensionReductionPreprocessor::apply_to_feature_matrix(CFeatures* features)
{
if (m_converter)
{
m_converter->set_target_dim(m_target_dim);
CDenseFeatures<float64_t>* embedding = m_converter->embed(features);
SGMatrix<float64_t> embedding_feature_matrix = embedding->steal_feature_matrix();
((CDenseFeatures<float64_t>*)features)->set_feature_matrix(embedding_feature_matrix);
delete embedding;
return embedding_feature_matrix;
}
else
{
SG_WARNING("Converter to process was not set.\n")
return ((CDenseFeatures<float64_t>*)features)->get_feature_matrix();
}
}
template<class ST> float64_t CDenseFeatures<ST>::dot(int32_t vec_idx1, CDotFeatures* df,
int32_t vec_idx2)
{
ASSERT(df);
ASSERT(df->get_feature_type() == get_feature_type());
ASSERT(df->get_feature_class() == get_feature_class());
CDenseFeatures<ST>* sf = (CDenseFeatures<ST>*) df;
int32_t len1, len2;
bool free1, free2;
ST* vec1 = get_feature_vector(vec_idx1, len1, free1);
ST* vec2 = sf->get_feature_vector(vec_idx2, len2, free2);
float64_t result = SGVector<ST>::dot(vec1, vec2, len1);
free_feature_vector(vec1, vec_idx1, free1);
sf->free_feature_vector(vec2, vec_idx2, free2);
return result;
}
int main(int argc,char *argv[])
{
init_shogun(&print_message,&print_message,&print_message);//initialising shogun without giving arguments shogun wont be able to print
int32_t x_n=4,x_d=2;//X dimensions : x_n for no of datapoints and x_d for dimensionality of data
SGMatrix<float64_t> fmatrix(x_d,x_n);
SG_SPRINT("\nTEST 1:\n\n");
/*Initialising Feature Matrix */
for (int i=0; i<x_n*x_d; i++)
fmatrix.matrix[i] = i+1;
SG_SPRINT("FEATURE MATRIX :\n");
CMath::display_matrix(fmatrix.matrix,x_d,x_n);
CDenseFeatures<float64_t>* features = new CDenseFeatures<float64_t>(fmatrix);
SG_REF(features);
/*Creating random labels */
CLabels* labels=new CLabels(x_n);
// create labels, two classes
labels->set_label(0,1);
labels->set_label(1,-1);
labels->set_label(2,1);
labels->set_label(3,1);
SG_REF(labels);
/*Working with Newton SVM */
float64_t lambda=1.0;
int32_t iter=20;
CNewtonSVM *nsvm = new CNewtonSVM(lambda,features,labels,iter);
SG_REF(nsvm);
nsvm->train();
SG_UNREF(labels);
SG_UNREF(nsvm);
SG_SPRINT("TEST 2:\n\n");
x_n=5;
x_d=3;
SGMatrix<float64_t> fmatrix2(x_d,x_n);
for (int i=0; i<x_n*x_d; i++)
fmatrix2.matrix[i] = i+1;
SG_SPRINT("FEATURE MATRIX :\n");
CMath::display_matrix(fmatrix2.matrix,x_d,x_n);
features->set_feature_matrix(fmatrix2);
SG_REF(features);
/*Creating random labels */
CLabels* labels2=new CLabels(x_n);
// create labels, two classes
labels2->set_label(0,1);
labels2->set_label(1,-1);
labels2->set_label(2,1);
labels2->set_label(3,1);
labels2->set_label(4,-1);
SG_REF(labels2);
/*Working with Newton SVM */
lambda=1.0;
iter=20;
CNewtonSVM *nsvm2 = new CNewtonSVM(lambda,features,labels2,iter);
SG_REF(nsvm2);
nsvm2->train();
SG_UNREF(labels2);
SG_UNREF(nsvm2);
SG_UNREF(features);
exit_shogun();
return 0;
}
void CLMNN::train(SGMatrix<float64_t> init_transform)
{
SG_DEBUG("Entering CLMNN::train().\n")
// Check training data and arguments, initializing, if necessary, init_transform
CLMNNImpl::check_training_setup(m_features, m_labels, init_transform);
// Initializations
// cast is safe, check_training_setup ensures features are dense
CDenseFeatures<float64_t>* x = static_cast<CDenseFeatures<float64_t>*>(m_features);
CMulticlassLabels* y = CLabelsFactory::to_multiclass(m_labels);
SG_DEBUG("%d input vectors with %d dimensions.\n", x->get_num_vectors(), x->get_num_features());
// Use Eigen matrix for the linear transform L. The Mahalanobis distance is L^T*L
MatrixXd L = Map<MatrixXd>(init_transform.matrix, init_transform.num_rows,
init_transform.num_cols);
// Compute target or genuine neighbours
SG_DEBUG("Finding target nearest neighbors.\n")
SGMatrix<index_t> target_nn = CLMNNImpl::find_target_nn(x, y, m_k);
// Initialize (sub-)gradient
SG_DEBUG("Summing outer products for (sub-)gradient initialization.\n")
MatrixXd gradient = (1-m_regularization)*CLMNNImpl::sum_outer_products(x, target_nn);
// Value of the objective function at every iteration
SGVector<float64_t> obj(m_maxiter);
// The step size is modified depending on how the objective changes, leave the
// step size member unchanged and use a local one
float64_t stepsize = m_stepsize;
// Last active set of impostors computed exactly, current and previous impostors sets
ImpostorsSetType exact_impostors, cur_impostors, prev_impostors;
// Iteration counter
uint32_t iter = 0;
// Criterion for termination
bool stop = false;
// Make space for the training statistics
m_statistics->resize(m_maxiter);
// Main loop
while (!stop)
{
SG_PROGRESS(iter, 0, m_maxiter)
// Find current set of impostors
SG_DEBUG("Finding impostors.\n")
cur_impostors = CLMNNImpl::find_impostors(x,y,L,target_nn,iter,m_correction);
SG_DEBUG("Found %d impostors in the current set.\n", cur_impostors.size())
// (Sub-) gradient computation
SG_DEBUG("Updating gradient.\n")
CLMNNImpl::update_gradient(x, gradient, cur_impostors, prev_impostors, m_regularization);
// Take gradient step
SG_DEBUG("Taking gradient step.\n")
CLMNNImpl::gradient_step(L, gradient, stepsize, m_diagonal);
// Compute the objective, trace of Mahalanobis distance matrix (L squared) times the gradient
// plus the number of current impostors to account for the margin
SG_DEBUG("Computing objective.\n")
obj[iter] = TRACE(L.transpose()*L,gradient) + m_regularization*cur_impostors.size();
// Correct step size
CLMNNImpl::correct_stepsize(stepsize, obj, iter);
// Check termination criterion
stop = CLMNNImpl::check_termination(stepsize, obj, iter, m_maxiter, m_stepsize_threshold, m_obj_threshold);
// Update iteration counter
iter = iter + 1;
// Update previous set of impostors
prev_impostors = cur_impostors;
// Store statistics for this iteration
m_statistics->set(iter-1, obj[iter-1], stepsize, cur_impostors.size());
SG_DEBUG("iteration=%d, objective=%.4f, #impostors=%4d, stepsize=%.4E\n",
iter, obj[iter-1], cur_impostors.size(), stepsize)
}
// Truncate statistics in case convergence was reached in less than maxiter
m_statistics->resize(iter);
// Store the transformation found in the class attribute
int32_t nfeats = x->get_num_features();
float64_t* cloned_data = SGMatrix<float64_t>::clone_matrix(L.data(), nfeats, nfeats);
m_linear_transform = SGMatrix<float64_t>(cloned_data, nfeats, nfeats);
SG_DEBUG("Leaving CLMNN::train().\n")
}
void CKMeans::clustknb(bool use_old_mus, float64_t *mus_start)
{
ASSERT(distance && distance->get_feature_type()==F_DREAL);
CDenseFeatures<float64_t>* lhs = (CDenseFeatures<float64_t>*) distance->get_lhs();
ASSERT(lhs && lhs->get_num_features()>0 && lhs->get_num_vectors()>0);
int32_t XSize=lhs->get_num_vectors();
dimensions=lhs->get_num_features();
int32_t i, changed=1;
const int32_t XDimk=dimensions*k;
int32_t iter=0;
R=SGVector<float64_t>(k);
mus=SGMatrix<float64_t>(dimensions, k);
int32_t *ClList=SG_CALLOC(int32_t, XSize);
float64_t *weights_set=SG_CALLOC(float64_t, k);
float64_t *dists=SG_CALLOC(float64_t, k*XSize);
///replace rhs feature vectors
CDenseFeatures<float64_t>* rhs_mus = new CDenseFeatures<float64_t>(0);
CFeatures* rhs_cache = distance->replace_rhs(rhs_mus);
int32_t vlen=0;
bool vfree=false;
float64_t* vec=NULL;
/* ClList=zeros(XSize,1) ; */
memset(ClList, 0, sizeof(int32_t)*XSize);
/* weights_set=zeros(k,1) ; */
memset(weights_set, 0, sizeof(float64_t)*k);
/* cluster_centers=zeros(dimensions, k) ; */
memset(mus.matrix, 0, sizeof(float64_t)*XDimk);
if (!use_old_mus)
{
for (i=0; i<XSize; i++)
{
const int32_t Cl=CMath::random(0, k-1);
int32_t j;
float64_t weight=Weights.vector[i];
weights_set[Cl]+=weight;
ClList[i]=Cl;
vec=lhs->get_feature_vector(i, vlen, vfree);
for (j=0; j<dimensions; j++)
mus.matrix[Cl*dimensions+j] += weight*vec[j];
lhs->free_feature_vector(vec, i, vfree);
}
for (i=0; i<k; i++)
{
int32_t j;
if (weights_set[i]!=0.0)
for (j=0; j<dimensions; j++)
mus.matrix[i*dimensions+j] /= weights_set[i];
}
}
else
{
ASSERT(mus_start);
/// set rhs to mus_start
rhs_mus->copy_feature_matrix(SGMatrix<float64_t>(mus_start,dimensions,k));
float64_t* p_dists=dists;
for(int32_t idx=0;idx<XSize;idx++,p_dists+=k)
distances_rhs(p_dists,0,k,idx);
p_dists=NULL;
for (i=0; i<XSize; i++)
{
float64_t mini=dists[i*k];
int32_t Cl = 0, j;
for (j=1; j<k; j++)
{
if (dists[i*k+j]<mini)
{
Cl=j;
mini=dists[i*k+j];
}
}
ClList[i]=Cl;
}
/* Compute the sum of all points belonging to a cluster
* and count the points */
for (i=0; i<XSize; i++)
{
const int32_t Cl = ClList[i];
float64_t weight=Weights.vector[i];
weights_set[Cl]+=weight;
#ifndef MUSRECALC
vec=lhs->get_feature_vector(i, vlen, vfree);
for (j=0; j<dimensions; j++)
mus.matrix[Cl*dimensions+j] += weight*vec[j];
lhs->free_feature_vector(vec, i, vfree);
#endif
}
#ifndef MUSRECALC
/* normalization to get the mean */
for (i=0; i<k; i++)
{
if (weights_set[i]!=0.0)
{
int32_t j;
for (j=0; j<dimensions; j++)
mus.matrix[i*dimensions+j] /= weights_set[i];
}
}
#endif
}
while (changed && (iter<max_iter))
{
iter++;
if (iter==max_iter-1)
SG_WARNING("kmeans clustering changed throughout %d iterations stopping...\n", max_iter-1);
if (iter%1000 == 0)
SG_INFO("Iteration[%d/%d]: Assignment of %i patterns changed.\n", iter, max_iter, changed);
changed=0;
#ifdef MUSRECALC
/* mus=zeros(dimensions, k) ; */
memset(mus.matrix, 0, sizeof(float64_t)*XDimk);
for (i=0; i<XSize; i++)
{
int32_t j;
int32_t Cl=ClList[i];
float64_t weight=Weights.vector[i];
vec=lhs->get_feature_vector(i, vlen, vfree);
for (j=0; j<dimensions; j++)
mus.matrix[Cl*dimensions+j] += weight*vec[j];
lhs->free_feature_vector(vec, i, vfree);
}
for (i=0; i<k; i++)
{
int32_t j;
if (weights_set[i]!=0.0)
for (j=0; j<dimensions; j++)
mus.matrix[i*dimensions+j] /= weights_set[i];
}
#endif
///update rhs
rhs_mus->copy_feature_matrix(mus);
for (i=0; i<XSize; i++)
{
/* ks=ceil(rand(1,XSize)*XSize) ; */
const int32_t Pat= CMath::random(0, XSize-1);
const int32_t ClList_Pat=ClList[Pat];
int32_t imini, j;
float64_t mini, weight;
weight=Weights.vector[Pat];
/* compute the distance of this point to all centers */
for(int32_t idx_k=0;idx_k<k;idx_k++)
dists[idx_k]=distance->distance(Pat,idx_k);
/* [mini,imini]=min(dists(:,i)) ; */
imini=0 ; mini=dists[0];
for (j=1; j<k; j++)
if (dists[j]<mini)
{
mini=dists[j];
imini=j;
}
if (imini!=ClList_Pat)
{
changed= changed + 1;
/* weights_set(imini) = weights_set(imini) + weight ; */
weights_set[imini]+= weight;
/* weights_set(j) = weights_set(j) - weight ; */
weights_set[ClList_Pat]-= weight;
vec=lhs->get_feature_vector(Pat, vlen, vfree);
for (j=0; j<dimensions; j++)
{
mus.matrix[imini*dimensions+j]-=(vec[j]
-mus.matrix[imini*dimensions+j])
*(weight/weights_set[imini]);
}
lhs->free_feature_vector(vec, Pat, vfree);
/* mu_new = mu_old - (x - mu_old)/(n-1) */
/* if weights_set(j)~=0 */
if (weights_set[ClList_Pat]!=0.0)
{
vec=lhs->get_feature_vector(Pat, vlen, vfree);
for (j=0; j<dimensions; j++)
{
mus.matrix[ClList_Pat*dimensions+j]-=
(vec[j]
-mus.matrix[ClList_Pat
*dimensions+j])
*(weight/weights_set[ClList_Pat]);
}
lhs->free_feature_vector(vec, Pat, vfree);
}
else
/* mus(:,j)=zeros(dimensions,1) ; */
for (j=0; j<dimensions; j++)
mus.matrix[ClList_Pat*dimensions+j]=0;
/* ClList(i)= imini ; */
ClList[Pat] = imini;
}
}
}
/* compute the ,,variances'' of the clusters */
for (i=0; i<k; i++)
{
float64_t rmin1=0;
float64_t rmin2=0;
bool first_round=true;
for (int32_t j=0; j<k; j++)
{
if (j!=i)
{
int32_t l;
float64_t dist = 0;
for (l=0; l<dimensions; l++)
{
dist+=CMath::sq(
mus.matrix[i*dimensions+l]
-mus.matrix[j*dimensions+l]);
}
if (first_round)
{
rmin1=dist;
rmin2=dist;
first_round=false;
}
else
{
if ((dist<rmin2) && (dist>=rmin1))
rmin2=dist;
if (dist<rmin1)
{
rmin2=rmin1;
rmin1=dist;
}
}
}
}
R.vector[i]=(0.7*CMath::sqrt(rmin1)+0.3*CMath::sqrt(rmin2));
}
distance->replace_rhs(rhs_cache);
delete rhs_mus;
SG_FREE(ClList);
SG_FREE(weights_set);
SG_FREE(dists);
SG_UNREF(lhs);
}
int main(int, char*[])
{
init_shogun_with_defaults();
#ifdef HAVE_LAPACK // for CDataGenerator::generate_gaussian()
// initialize the random number generator with a fixed seed, for repeatability
CMath::init_random(10);
// Prepare the training data
const int num_features = 20;
const int num_classes = 4;
const int num_examples_per_class = 20;
SGMatrix<float64_t> X;
try
{
X = CDataGenerator::generate_gaussians(
num_examples_per_class,num_classes,num_features);
}
catch (ShogunException e)
{
// out of memory
SG_SPRINT(e.get_exception_string());
return 0;
}
CDenseFeatures<float64_t>* features = new CDenseFeatures<float64_t>(X);
// Create a deep autoencoder
CNeuralLayers* layers = new CNeuralLayers();
layers
->input(num_features)
->rectified_linear(10)->rectified_linear(5)->rectified_linear(10)
->linear(num_features);
CDeepAutoencoder* ae = new CDeepAutoencoder(layers->done());
// uncomment this line to enable info logging
// ae->io->set_loglevel(MSG_INFO);
// pre-train
ae->pt_epsilon.set_const(1e-6);
ae->pre_train(features);
// fine-tune
ae->train(features);
// reconstruct the data
CDenseFeatures<float64_t>* reconstructions = ae->reconstruct(features);
SGMatrix<float64_t> X_reconstructed = reconstructions->get_feature_matrix();
// find the average difference between the data and the reconstructions
float64_t avg_diff = 0;
int32_t N = X.num_rows*X.num_cols;
for (int32_t i=0; i<N; i++)
avg_diff += CMath::abs(X[i]-X_reconstructed[i])/CMath::abs(X[i]);
avg_diff /= N;
SG_SINFO("Average difference = %f %\n", avg_diff*100);
// Clean up
SG_UNREF(ae);
SG_UNREF(layers);
SG_UNREF(features);
SG_UNREF(reconstructions);
#endif
exit_shogun();
return 0;
}
CLabels* CAUCKernel::setup_auc_maximization(CLabels* labels)
{
SG_INFO( "setting up AUC maximization\n") ;
ASSERT(labels);
ASSERT(labels->get_label_type() == LT_BINARY);
labels->ensure_valid();
// get the original labels
SGVector<int32_t> int_labels=((CBinaryLabels*) labels)->get_int_labels();
ASSERT(subkernel->get_num_vec_rhs()==int_labels.vlen);
// count positive and negative
int32_t num_pos=0;
int32_t num_neg=0;
for (int32_t i=0; i<int_labels.vlen; i++)
{
if (int_labels.vector[i]==1)
num_pos++;
else
num_neg++;
}
// create AUC features and labels (alternate labels)
int32_t num_auc = num_pos*num_neg;
SG_INFO("num_pos: %i num_neg: %i num_auc: %i\n", num_pos, num_neg, num_auc);
SGMatrix<uint16_t> features_auc(2,num_auc);
int32_t* labels_auc = SG_MALLOC(int32_t, num_auc);
int32_t n=0 ;
for (int32_t i=0; i<int_labels.vlen; i++)
{
if (int_labels.vector[i]!=1)
continue;
for (int32_t j=0; j<int_labels.vlen; j++)
{
if (int_labels.vector[j]!=-1)
continue;
// create about as many positively as negatively labeled examples
if (n%2==0)
{
features_auc.matrix[n*2]=i;
features_auc.matrix[n*2+1]=j;
labels_auc[n]=1;
}
else
{
features_auc.matrix[n*2]=j;
features_auc.matrix[n*2+1]=i;
labels_auc[n]=-1;
}
n++;
ASSERT(n<=num_auc);
}
}
// create label object and attach it to svm
CBinaryLabels* lab_auc = new CBinaryLabels(num_auc);
lab_auc->set_int_labels(SGVector<int32_t>(labels_auc, num_auc, false));
SG_REF(lab_auc);
// create feature object
CDenseFeatures<uint16_t>* f = new CDenseFeatures<uint16_t>(0);
f->set_feature_matrix(features_auc);
// create AUC kernel and attach the features
init(f,f);
SG_FREE(labels_auc);
return lab_auc;
}
