apaam/mlpack
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

runs but there is an instability in

Browse Source 
ComputeLLE
This commit is contained in:
vasiloglou
2008-01-18 17:19:31 +00:00
parent b1607d1246
commit 80d71f2e2b
1 changed files with 37 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastlib/u/nvasil/kernel_pca/kernel_pca_impl.h
+37 -19
View File
@@ -118,7 +118,10 @@ void KernelPCA::EstimateBandwidth(double *bandwidth) {
}
*bandwidth=mean/count;
}
// It is not always know if the k nearest neighbors includes the same point with
// 0 distance. It is highly likely in cases where the query tree and the reference tree
// come from different structures that refer to the same dataset. We take care about
// that in this function.
void KernelPCA::ComputeLLE(index_t num_of_eigenvalues,
Matrix *eigen_vectors,
std::vector<double> *eigen_values) {
@@ -132,37 +135,52 @@ void KernelPCA::ComputeLLE(index_t num_of_eigenvalues,
Vector point;
point.Init(dimension_);
Matrix neighbor_vals;
neighbor_vals.Init(dimension_, knns_-1);
Matrix covariance(knns_-1, knns_-1);
// We initialize everything with knns_ although it is highly likely that
// if the k nearest neighbors include the same point then we will need
// a smaller vector knns_-1
neighbor_vals.Init(dimension_, knns_);
Matrix covariance(knns_, knns_);
Matrix t_covariance(knns_-1, knns_-1);
Vector ones;
ones.Init(knns_-1);
ones.Init(knns_);
ones.SetAll(1);
Vector weights;
index_t neighbors[knns_];
index_t i;
kernel_matrix_.Init(data_.n_rows(),
data_.n_rows(),
kernel_matrix_.Init(data_.n_cols(),
data_.n_cols(),
knns_);
last_point=0;
while (!feof(fp)) {
fscanf(fp, "%llu %llu %lg", &p1, &p2, &dist);
fscanf(fp, "%llu %llu %lg\n", &p1, &p2, &dist);
if (dist==0) {
continue;
}
i=0;
if (p1==last_point) {
memcpy(neighbor_vals.GetColumnPtr(i), data_.GetColumnPtr(p2),
sizeof(double)*dimension_);
neighbors[i]=p2;
la::SubFrom(dimension_, point.ptr(), neighbor_vals.GetColumnPtr(i));
} else {
if (p1!=last_point) {
point.CopyValues(data_.GetColumnPtr(p1));
last_point=p1;
i=0;
la::MulTransAOverwrite(neighbor_vals, neighbor_vals, &covariance);
la::SolveInit(covariance, ones, &weights);
kernel_matrix_.LoadRow(p1, knns_-1, neighbors, weights.ptr());
weights.Destruct();
if (i==knns_) {
la::MulTransAOverwrite(neighbor_vals, neighbor_vals, &covariance);
la::SolveInit(covariance, ones, &weights);
kernel_matrix_.LoadRow(p1, knns_, neighbors, weights.ptr());
weights.Destruct();
} else {
Vector t_ones;
Matrix t_neighbor_vals;
Vector t_weights;
t_ones.Alias(ones.ptr(), i);
t_neighbor_vals.Alias(neighbor_vals.ptr(), dimension_, i);
la::MulTransAOverwrite(t_neighbor_vals, t_neighbor_vals, &t_covariance);
la::SolveInit(t_covariance, t_ones, &t_weights);
kernel_matrix_.LoadRow(p1, i, neighbors, t_weights.ptr());
}
i=0;
}
memcpy(neighbor_vals.GetColumnPtr(i), data_.GetColumnPtr(p2),
sizeof(double)*dimension_);
neighbors[i]=p2;
la::SubFrom(dimension_, point.ptr(), neighbor_vals.GetColumnPtr(i));
i++;
}
kernel_matrix_.Negate();
kernel_matrix_.SetDiagonal(1.0);