increasing number of components

[Rajalakshmi-Palaniappan]

Dear Seung lab,

I have this weird issue. I have around 25k labels obtained through connected component labelling. After skeletonization using kimimaro with the following parameters, I can see in some skeletons (swc) there are two root nodes(nodes with -1 as parent). I checked and tried different connectivity in the connected component labelling algorithm and still the issue persists. It would be very helpful if you could point out the reason for this happening

skels = kimimaro.skeletonize(
Labels,
teasar_params={
'scale': 1.1,
'const': 10, # set scale = 1.1 and const = 10 for the data with blood vessels
'pdrf_exponent': 4,
'pdrf_scale': 100000,
'soma_detection_threshold': 10, # physical units
'soma_acceptance_threshold': 10, # physical units
'soma_invalidation_scale': 1.0,
'soma_invalidation_const': 10, # physical units
'max_paths': None, # default None
},
dust_threshold=0, # skip connected components with fewer than this many voxels
anisotropy=(1,1,1), # default True
fix_branching=True, # default True
fix_borders=True, # default True
progress=True, # default False, show progress bar
parallel=1, # <= 0 all cpu, 1 single process, 2+ multiprocess
parallel_chunk_size=50, # how many skeletons to process before updating progress bar
)

[william-silversmith]

This is not strange if some of the labels appear twice, but it sounds like you already checked that. If you have a data sample to share, it might be possible for me to see what's going on. Sorry about the delay, I just saw this issue!