Skip to contents

Helper function to generate default parameters for the fast clustering for the doublet detection methods

params_fast_cluster_default.Rd

Helper function to generate default parameters for the fast clustering for the doublet detection methods

Usage

params_fast_cluster_default()

Value

A named list with the following elements:

  • km_type - String. The type of k-means clustering. One of c("minibatch", "standard"). Defaults to "minibatch".

  • n_centroids - Integer or NULL. The number of centroids to use. NULL uses sqrt(N_cells) * 4 centroids. Defaults to NULL.

  • kmeans_iters - Integer. Number of maximum k-means iterations. Defaults to 100L.

  • batch_size - Integer. Max batch size, capped at N_cells / 2 depending on the data set. Defaults to 4098L.