dot-preprocessData: Preprocess data

In jbergenstrahle/SpatialCPie: Cluster analysis of Spatial Transcriptomics data

Description

Preprocesses input data for .makeServer.

Usage

.preprocessData(counts, margin, resolutions, assignmentFunction,
  coordinates = NULL)

Arguments

counts	count matrix. rownames should correspond to genes and colnames should correspond to spot coordinates.
margin	which margin of the count matrix to cluster. Valid values are c("spot", "sample", "gene", "feature").
resolutions	vector of resolutions to cluster.
assignmentFunction	function to compute cluster assignments. The function should have the following signature: integer (number of clusters) -> (m, n) feature matrix -> m-length vector (cluster assignment of each data point).
coordinates	optional data.frame with pixel coordinates for each spot. rownames should correspond to the colnames of counts and the columns x and y should specify the pixel coordinates of the spots.

Value

list with the following elements:

• $assignments: tidy assignments

• $means: cluster means

• $scores: cluster scores for each spot in each resolution

• $colors: cluster colors

• $coordinates: spot coordinates, either from coordinates or parsed from assignments

• $featureName: name of the clustered feature (the "opposite" of margin)

jbergenstrahle/SpatialCPie documentation built on April 5, 2020, 7:56 a.m.