This function can be used to exclude from the inference step those segments that are unimodal.

The function requires and returns an (R)CONGAS+ object.

segments_selector_congas(
  obj,
  multiome = F,
  K_max = 3,
  score = "BIC",
  lambda = 0.5,
  cores_ratio = 0.5,
  CUDA = F,
  binom_limits = c(40, 1000)
)

Arguments

obj

An rcongasplus object.

K_max

Max Number of clusters to test for the congas runs on single segments

score

Model selection score to be used to select the number of clusters

lambda

lambda to use for each congas run

cores_ratio

fraction of cores that we be used to run the single CONGAS+ inferences in parallel.

Value

The object obj where segments have been identified and removed.