Cell-cell Interaction Fails if Any Sample Lacks Any Cell Type

[DarioS]

Some samples may only have a subset of cell types. For example, c.names has 5 levels but cluster has 4 distinct numeric values.

Browse[2]> celltype <- as.factor(alldata$celltype[alldata$sample == x]) # Code from helper_CCI in helper_CCI.R
Browse[2]> table(celltype)
celltype
 dCAF  iCAF  mCAF tpCAF  vCAF 
  202   492   242   737     0
Browse[2]> celltype_numeric <- as.numeric(celltype)  # Code from helper_CCI in helper_CCI.R
Browse[2]> table(celltype_numeric)
celltype_numeric
  1   2   3   4 
202 492 242 737 
Browse[2]> signal <- cell_signaling(data = data_dataframe,  genes = rownames(data_dataframe), cluster = celltype_numeric,
                                    c.names = levels(celltype), write = FALSE)  # Code from helper_CCI in helper_CCI.R
Error in SingleCellSignalR::cell_signaling(data = data_dataframe, genes = rownames(data_dataframe),  : 
  The length of c.names must be equal to the number of clusters and must contain no duplicates.

scFeatures should gracefully drop empty factor levels like aFactor <- droplevels(aFactor) before cell_signaling function call.