Xmipp
v3.23.11-Nereus
|
#include <dimred_tools.h>
Public Member Functions | |
void | generateNewDataset (const DatasetType &type, int N=1000, double noise=0.05) |
Public Attributes | |
Matrix2D< double > | X |
Matrix2D< double > | t |
Matrix1D< unsigned char > | label |
Definition at line 43 of file dimred_tools.h.
void GenerateData::generateNewDataset | ( | const DatasetType & | type, |
int | N = 1000 , |
||
double | noise = 0.05 |
||
) |
Generate data with a given number of points and a given method. Generates an artificial dataset. Possible datasets are: 'swiss' for the Swiss roll dataset, 'helix' for the helix dataset, 'twinpeaks' for the twinpeaks dataset, '3d_clusters' for the 3D clusters dataset, and 'intersect' for the intersecting dataset. The variable n indicates the number of datapoints to generate. The variable noise indicates the amount of noise that is added to the data. The function generates the high-dimensional dataset in X, and corresponding labels in labels. In addition, the function keeps the coordinates of the datapoints on the underlying manifold in t.
Definition at line 31 of file dimred_tools.cpp.
Matrix1D<unsigned char> GenerateData::label |
Vector of labels for the observations
Definition at line 56 of file dimred_tools.h.
Matrix2D<double> GenerateData::t |
Underlying manifold coordinates. Eeach row of the matrix corresponds to the manifold coordinates of the observation in X.
Definition at line 53 of file dimred_tools.h.
Matrix2D<double> GenerateData::X |
Generated data. Each row of the matrix is an individual observation
Definition at line 48 of file dimred_tools.h.