When attempting sub-clustering the function subcluster_labels has a minimum number of clusters argument which is by default set to 50. However, when we have cases where the number of cells in a broad cluster do not meet this minimum, the function stops and presents and error. This should instead present a warning and then if the ncells is below the set argument, the subclustering can be skipped and the all the cells assigned to the one cluster - also need a way to handle this downstream. I think this logic is already supported but will need to double check this.
When attempting sub-clustering the function subcluster_labels has a minimum number of clusters argument which is by default set to 50. However, when we have cases where the number of cells in a broad cluster do not meet this minimum, the function stops and presents and error. This should instead present a warning and then if the ncells is below the set argument, the subclustering can be skipped and the all the cells assigned to the one cluster - also need a way to handle this downstream. I think this logic is already supported but will need to double check this.