clusterJP: Jarvis-Patrick Clustering
In wind22zhu/BioMedR: Generating Various Molecular Representations for Chemicals, Proteins, DNAs, RNAs and Their Interactions

Description Usage Arguments Details Value Author(s) References See Also Examples

Jarvis-Patrick Clustering

1	clusterJP(nnm, k, mode = "a1a2b", linkage = "single")

`nnm`	A nearest neighbor table, as produced by `nearestNeighbors`.
`k`	Minimum number of nearest neighbors two rows (items) in the nearest neighbor table need to have in common to join them into the same cluster.
`mode`	If `mode = "a1a2b"` (default), the clustering is run with both requirements (a) and (b); if `mode = "a1b"` then (a) is relaxed to a unidirectional requirement; and if `mode = "b"` then only requirement (b) is used. The size of the clusters generated by the different methods increases in this order: "a1a2b" < "a1b" < "b". The run time of method "a1a2b" follows a close to linear relationship, while it is nearly quadratic for the much more exhaustive method "b". Only methods "a1a2b" and "a1b" are suitable for clustering very large data sets (e.g. >50,000 items) in a reasonable amount of time.
`linkage`	Can be one of "single", "average", or "complete", for single linkage, average linkage and complete linkage merge requirements, respectively. In the context of Jarvis-Patrick, average linkage means that at least half of the pairs between the clusters under consideration must pass the merge requirement. Similarly, for complete linkage, all pairs must pass the merge requirement. Single linkage is the normal case for Jarvis-Patrick and just means that at least one pair must meet the requirement.

Function to perform Jarvis-Patrick clustering. The algorithm requires a nearest neighbor table, which consists of neighbors for each item in the dataset. This information is then used to join items into clusters with the following requirements: (a) they are contained in each other's neighbor list (b) they share at least k nearest neighbors The nearest neighbor table can be computed with NNeighbors. For standard Jarvis-Patrick clustering, this function takes the number of neighbors to keep for each item.

Depending on the setting under the type argument, the function returns the clustering result in a named vector or a nearest neighbor table as matrix.

Min-feng Zhu <wind2zhu@163.com>

Jarvis RA, Patrick EA (1973) Clustering Using a Similarity Measure Based on Shared Near Neighbors. IEEE Transactions on Computers, C22, 1025-1034. URLs: http://davide.eynard.it/teaching/2012_PAMI/JP.pdf, http://www.btluke.com/jpclust.html, http://www.daylight.com/dayhtml/doc/cluster/index.pdf

see NNeighbors for nearest neighbors

data(sdfbcl)
apbcl <- convSDFtoAP(sdfbcl)
fpbcl <- convAPtoFP(apbcl)
clusterJP(NNeighbors(apbcl, numNbrs = 6), k = 5, mode = "a1a2b")
clusterJP(NNeighbors(fpbcl, numNbrs = 6), k = 5, mode = "a1b")
clusterJP(NNeighbors(apbcl, cutoff = 0.6), k = 2, mode = 'b')