fitModal | R Documentation |
Fit a bimodal gaussian distribution to a set of observations.
fitModal(
x,
m,
prob = 0.95,
coverage = 0.8,
size = 10,
assign = FALSE,
boolean = FALSE,
verbose = TRUE,
maxit = 5000,
maxrestarts = 100,
bySampling = FALSE,
nsamp = 200,
...
)
x |
a named numeric vector of cells/observations or a matrix of genes X cells (variables X observations). If the latter, the column means are first computed. |
m |
number of components (modes). Default: 2 |
prob |
a numeric value >= 0 and <= 1; the minimum posterior probability required for an observation to be assigned to a mode. Default: 0.95 |
coverage |
the fraction of observations that must have a posterior probability higher than <prob> to one of two modes in order for the distribution to qualify as bimodal. Default: 0.8 |
size |
the minimum number of observations that must be assigned to a mode in order for the distribution to qualify as bimodal. Default: 10 |
assign |
if set to TRUE, returns a list of length two containing the vector names that were assigned to each mode. Default: FALSE |
boolean |
if set to TRUE, returns a boolean value indicating whether the distribution is bimodal. Default: FALSE |
verbose |
print progress messages. Default: TRUE |
maxit |
the maximum number of iterations. Default: 5000 |
maxrestarts |
the maximum number of restarts allowed. See |
The posterior probabilities of each observation to one of two modes. If boolean = TRUE, return a boolean value indicating whether bimodality was found. If assign = TRUE, return a list of length two with the observations (IDs) in each mode.
normalmixEM
cna = infercna(m = useData(), refCells = refCells)
# Malignant cells only (remove columns corresponding to refCells)
cna = cna[, !colnames(cna) %in% unlist(refCells)]
cnaByChr = splitGenes(cna, by = 'chr')
sapply(cnaByChr, fitBimodal, assign = TRUE)
sapply(cnaByChr, fitBimodal, boolean = TRUE)
sapply(cnaByChr, fitBimodal, boolean = TRUE, coverage = 0.5)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.