cluster_split.RdSplit clusters into new clusters based on temperature for the deterministic annealing algorithm
cluster_split(pca.components, parameter.estimates, clusters, temperature, decay.step, max.clusters, num.cores)
| pca.components | Expects a matrix with n rows as the cells and m columns as the principal components. |
|---|---|
| parameter.estimates | Expects a list of parameters (mu, covariance, and weights) from initialize_gmm or maximization_step |
| clusters | Expect the most likely cluster assignments |
| temperature | Current temperature to use to split clusters |
| decay.step | Data.frame with the first column as the step number and the second column as the temperature for that step |
| max.clusters | Maximum number of clusters for the dataset |