Too many clusters generated when using default parameters
KshitijAggarwal opened this issue · comments
Running the scan with the default parameters of clustering is generating too many candidates, even if there was just one transient in the dataset (found when running pipeline on 16A-459_TEST_1hr_000.57633.66130137732.scan7.cut
), which implies that the clustering is not being done efficiently. Increasing the min_cluster_size
while setting the state resolves the issue though.
I'll bump the min_cluster_size
up by 1 to a value of 3. Hopefully that helps a bit.