Skip to contents

Generates the sNN graph for igraph

rs_sc_snn.Rd

[Experimental] This function takes a kNN matrix and generates the inputs for an SNN graph based on it.

Usage

rs_sc_snn(knn_mat, snn_method, limited_graph, pruning, verbose)

Arguments

knn_mat

Integer matrix. Rows represent cells and the columns represent the neighbours (0-indexed!).

snn_method

String. Which method to use to calculate the similarity. Choice of c("jaccard", "rank"); anything else errors.

limited_graph

Boolean. Shall the sNNs only be calculated between direct neighbours in the graph, or between all possible combinations.

pruning

Float. Below which similarity value to prune the weight to 0.

verbose

Integer. 0L - quiet; 1L - normal verbosity; 2L - detailed verbosity.

Value

A list with the following items:

  • edges - Integer vector with the flattened sNN edge pairs (1-indexed!), ready for igraph::add_edges().

  • weights - sNN weights of the pairs above.