celda_G {celda} | R Documentation |
Clusters the rows of a count matrix containing single-cell data into L modules.
celda_G(counts, L, beta = 1, delta = 1, gamma = 1, stopIter = 10, maxIter = 200, splitOnIter = 10, splitOnLast = TRUE, nchains = 3, yInitialize = c("split", "random", "predefined"), countChecksum = NULL, yInit = NULL, logfile = NULL, verbose = TRUE)
counts |
Integer matrix. Rows represent features and columns represent cells. |
L |
Integer. Number of feature modules. |
beta |
Numeric. Concentration parameter for Phi. Adds a pseudocount to each feature module in each cell. Default 1. |
delta |
Numeric. Concentration parameter for Psi. Adds a pseudocount to each feature in each module. Default 1. |
gamma |
Numeric. Concentration parameter for Eta. Adds a pseudocount to the number of features in each module. Default 1. |
stopIter |
Integer. Number of iterations without improvement in the log likelihood to stop inference. Default 10. |
maxIter |
Integer. Maximum number of iterations of Gibbs sampling to perform. Default 200. |
splitOnIter |
Integer. On every 'splitOnIter' iteration, a heuristic will be applied to determine if a feature module should be reassigned and another feature module should be split into two clusters. To disable splitting, set to -1. Default 10. |
splitOnLast |
Integer. After 'stopIter' iterations have been performed without improvement, a heuristic will be applied to determine if a cell population should be reassigned and another cell population should be split into two clusters. If a split occurs, then 'stopIter' will be reset. Default TRUE. |
nchains |
Integer. Number of random cluster initializations. Default 3. |
yInitialize |
Chararacter. One of 'random', 'split', or 'predefined'. With 'random', features are randomly assigned to a modules. With 'split', features will be split into sqrt(L) modules and then each module will be subsequently split into another sqrt(L) modules. With 'predefined', values in ‘yInit' will be used to initialize 'y'. Default ’split'. |
countChecksum |
Character. An MD5 checksum for the 'counts' matrix. Default NULL. |
yInit |
Integer vector. Sets initial starting values of y. If NULL, starting values for each feature will be randomly sampled from '1:L'. ‘yInit' can only be used when 'initialize = ’random''. Default NULL. |
logfile |
Character. Messages will be redirected to a file named 'logfile'. If NULL, messages will be printed to stdout. Default NULL. |
verbose |
Logical. Whether to print log messages. Default TRUE. |
An object of class 'celda_G' with the feature module clusters stored in 'y'.
'celda_C()' for cell clustering and 'celda_CG()' for simultaneous clustering of features and cells. 'celdaGridSearch()' can be used to run multiple values of L and multiple chains in parallel.
data(celdaGSim) celdaMod <- celda_G(celdaGSim$counts, L = celdaGSim$L)