clusteringOptics: Perform Optics analysis of mapping.
In sdrowland/RSMod: Creates a 2D or 3D mapping of RNA-Seq Data

Description Usage Arguments Details Value Author(s) Examples

View source: R/clusteringOptics.R

Generates a 2D mapping dataframe with module selection using Optics from the dbscan package and KNNdist.

1	clusteringOptics(cluster, cutoff)

`cluster`	Data frame obtained from BH-SNE analysis. Requires x and y coordinates for each point in the 2D mapping.
`cutoff`	The neighborhood size cutoff value based on KNNdist analysis. The typical range for BH-SNE mappings is between 2 and 5, other mappings will vary. Typical "optimal" value falls along the linear portion of the graph, but this will vary by dataset and type.

An x/y coordiante mapping is required for clusteringOptics. The OPTICS portion of the dbscan package is used along with KNNdist to determine the cutoff points for clusters. The value supplied to cutoff provides a value to KNNdist to generate the optimal eps for dbscan.

Returns a matrix with the orginal x and y coordinates and module number assignments listed as colors. These values are used for plotting.

Steven D. Rowland