Closed EdwardAgboraw closed 3 years ago
Hi,
The voting procedure is already integrated in the package. You don't need to enable it by yourself.
If the memory is a concern for you, you can convert the input matrix to a sparse matrix, and set the parameter sparse=TRUE
for scDHA
function. This would greatly reduce the memory usage. However, the sparse matrix must not have any negative value. Let me know if you have any problem using the feature.
In the paper, a voting-procedure that applies the clustering algorithm on a subset of the data points is mentioned. Is this option possible using this package, and, if yes, how can it be accessed?