Utility class containing static hierarchical clustering methods. More...
#include <clusthierarchical.h>
Static Public Member Functions | |
| static CHierarchy * | Cluster (const CDistanceMatrix &MatSimilarities) |
| Hierarchically cluster a set of elements with the given pairwise similarity scores. | |
| static CHierarchy * | Cluster (const CDistanceMatrix &MatSimilarities, const std::vector< bool > &vecfIncluded) |
| Hierarchically cluster a subset of elements with the given pairwise similarity scores. | |
Detailed Description
Utility class containing static hierarchical clustering methods.
Definition at line 157 of file clusthierarchical.h.
Member Function Documentation
| CHierarchy * Sleipnir::CClustHierarchical::Cluster | ( | const CDistanceMatrix & | MatSimilarities | ) | [static] |
Hierarchically cluster a set of elements with the given pairwise similarity scores.
- Parameters:
-
MatSimilarities Matrix of similarity scores between each pair of elements.
- Returns:
- Resulting clustered hierarchy; null on failure.
Efficiently performs hierarchical clustering on a set of elements using a precalculated pairwise similarity matrix. This similarity matrix is often generated by applying some similarity measure to the genes in a PCL file. The leaves of the resulting hierarchy represent elements of the input matrix and are identified by their indices in this matrix. Clusters are joined using UPGMA (i.e. average linkage); the pair with the greatest similarity is joined first, then the next greatest, and so forth.
- See also:
- CClustKMeans::Cluster
Definition at line 206 of file clusthierarchical.cpp.
Referenced by Cluster().
| CHierarchy * Sleipnir::CClustHierarchical::Cluster | ( | const CDistanceMatrix & | MatSimilarities, |
| const std::vector< bool > & | vecfIncluded | ||
| ) | [static] |
Hierarchically cluster a subset of elements with the given pairwise similarity scores.
- Parameters:
-
MatSimilarities Matrix of similarity scores between each pair of elements. vecfIncluded Vector indicating which elements should be clustered (true to be included, false to be excluded).
- Returns:
- Resulting clustered hierarchy; null on failure.
Efficiently performs hierarchical clustering on a subset of elements using a precalculated pairwise similarity matrix. This similarity matrix is often generated by applying some similarity measure to the genes in a PCL file. The leaves of the resulting hierarchy represent elements of the input matrix and are identified by their indices in this matrix. Clusters are joined using UPGMA (i.e. average linkage); the pair with the greatest similarity is joined first, then the next greatest, and so forth. Indices for which vecfIncluded is false are ignored.
- Remarks:
- vecfIncluded must be of the same size as MatSimilarities.
- See also:
- CClustKMeans::Cluster
Definition at line 182 of file clusthierarchical.cpp.
References Cluster().
The documentation for this class was generated from the following files:
