DRAGONsc clustering algorithm

Takes input from, PCA components, temperature decay steps, maximum iterations, delta log likelihood for convergence at each step, maximum number of resulting clusters, and number of cores for parallelization. Has options for saving intermediate files and for verbose output.

dragonsc(pca.components, temp.decay.steps, max.iterations,
  delta.log.likelihood, max.clusters, save.intermediate.files = FALSE,
  num.cores, verbose = FALSE)

Arguments

pca.components	Expects a matrix with n rows as the cells and m columns as the principal components.
temp.decay.steps	Temperature progressed used for the algorithm. Expects a two column data frame, with columns named "step" and "temperature"
max.iterations	Maximum number of iterations of the expectation maximization at each temperature
delta.log.likelihood	Change in the log likelihood between steps of the expectation maximization for convergence
max.clusters	Maximum number of clusters in the final step of the algorithm
save.intermediate.files	Whether to save files at each temperature of the algorithm. Default is FALSE, but can be very informative to view intermediate results.
num.cores	Number of cores to use for the algorithm. Uses doParallel as a backend.
verbose	Whether to display the current step of the algorithm and the log likelihood for each iteration of the expectation maximization algorithm. Default is FALSE.