Failures for some containerized sorters, but not others

[vncntprvst]

I was testing some sorters that I typically don't use, and got some issues:

• HerdingSpikes

hs_output = sorting_dir / 'sorting_herdingspikes'
container_image = 'spikeinterface/herdingspikes-base'

params_hs = {}  # use SI defaults

sorting_hs = si.run_sorter(
    sorter_name='herdingspikes',
    recording=rec_saved,
    folder=hs_output,
    remove_existing_folder=True,
    docker_image=container_image, #True, 
    verbose=True,
    **params_hs
)

installation_mode='auto' switching to installation_mode: 'github'
Starting container
Installing spikeinterface with github in container
Installing neo with pypi in container
Running herdingspikes sorter inside spikeinterface/herdingspikes-base
Stopping container

---------------------------------------------------------------------------
SpikeSortingError                         Traceback (most recent call last)
Cell In[27], line 6
      2 container_image = 'spikeinterface/herdingspikes-base'
      4 params_hs = {}  # use SI defaults
----> 6 sorting_hs = si.run_sorter(
      7     sorter_name='herdingspikes',
      8     recording=rec,
      9     folder=hs_output,
     10     remove_existing_folder=True,
     11     docker_image=container_image, #True, 
     12     verbose=True,
     13     **params_hs
     14 )
     15 print('HerdingSpikes units:', len(sorting_hs.get_unit_ids()))

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:204, in run_sorter(sorter_name, recording, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, docker_image, singularity_image, delete_container_files, with_output, **sorter_params)
    198         if not has_spython():
    199             raise RuntimeError(
    200                 "The python `spython` package must be installed to "
    201                 "run singularity. Install with `pip install spython`"
    202             )
--> 204     return run_sorter_container(
    205         container_image=container_image,
    206         mode=mode,
    207         **common_kwargs,
    208     )
    210 return run_sorter_local(**common_kwargs)

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:681, in run_sorter_container(sorter_name, recording, mode, container_image, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, with_output, delete_container_files, extra_requirements, installation_mode, spikeinterface_version, spikeinterface_folder_source, **sorter_params)
    679 if run_error:
    680     if raise_error:
--> 681         raise SpikeSortingError(f"Spike sorting in {mode} failed with the following error:\n{run_sorter_output}")
    682 else:
    683     if with_output:

SpikeSortingError: Spike sorting in docker failed with the following error:

• Combinato

combinato_output = sorting_dir / 'sorting_combinato'
container_image = 'spikeinterface/combinato-base'

params_combinato = {}  # use SI defaults

sorting_combinato = si.run_sorter(
    sorter_name='combinato',
    recording=rec,
    folder=combinato_output,
    remove_existing_folder=True,
    docker_image=container_image, #True, 
    verbose=True,
    **params_combinato
)

installation_mode='auto' switching to installation_mode: 'github'
Starting container
Installing spikeinterface with github in container
Installing neo with pypi in container
Running combinato sorter inside spikeinterface/combinato-base
Stopping container

---------------------------------------------------------------------------
SpikeSortingError                         Traceback (most recent call last)
Cell In[26], line 6
      2 container_image = 'spikeinterface/combinato-base'
      4 params_combinato = {}  # use SI defaults
----> 6 sorting_combinato = si.run_sorter(
      7     sorter_name='combinato',
      8     recording=rec,
      9     folder=combinato_output,
     10     remove_existing_folder=True,
     11     docker_image=container_image, #True, 
     12     verbose=True,
     13     **params_combinato
     14 )
     15 print('Combinato units:', len(sorting_combinato.get_unit_ids()))

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:204, in run_sorter(sorter_name, recording, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, docker_image, singularity_image, delete_container_files, with_output, **sorter_params)
    198         if not has_spython():
    199             raise RuntimeError(
    200                 "The python `spython` package must be installed to "
    201                 "run singularity. Install with `pip install spython`"
    202             )
--> 204     return run_sorter_container(
    205         container_image=container_image,
    206         mode=mode,
    207         **common_kwargs,
    208     )
    210 return run_sorter_local(**common_kwargs)

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:681, in run_sorter_container(sorter_name, recording, mode, container_image, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, with_output, delete_container_files, extra_requirements, installation_mode, spikeinterface_version, spikeinterface_folder_source, **sorter_params)
    679 if run_error:
    680     if raise_error:
--> 681         raise SpikeSortingError(f"Spike sorting in {mode} failed with the following error:\n{run_sorter_output}")
    682 else:
    683     if with_output:

SpikeSortingError: Spike sorting in docker failed with the following error:
Warning! The recording is already filtered, but combinato filter is enabled
Traceback (most recent call last):
  File "/home/wanglab/data/TJ_Oscillator/LO_04/2025_08_26/LO_04_004/processed_data/spike_sorting_output_nb/in_container_sorter_script.py", line 23, in <module>
    sorting = run_sorter_local(
              ^^^^^^^^^^^^^^^^^
  File "/root/.local/lib/python3.12/site-packages/spikeinterface/sorters/runsorter.py", line 300, in run_sorter_local
    SorterClass.setup_recording(recording, folder, verbose=verbose)
  File "/root/.local/lib/python3.12/site-packages/spikeinterface/sorters/basesorter.py", line 242, in setup_recording
    cls._setup_recording(recording, sorter_output_folder, sorter_params, verbose)
  File "/root/.local/lib/python3.12/site-packages/spikeinterface/sorters/external/combinato.py", line 133, in _setup_recording
    assert len(chan_ids) == 1, "combinato is a single-channel recording"
           ^^^^^^^^^^^^^^^^^^
AssertionError: combinato is a single-channel recording

Sorting the same recording object worked with Kilosort4 and MountainSort5 (containers), and internal sorters Spyking Circus2 and Tridesclous2 (not containers).

Is it an issue with passing appropriate arguments to those two sorters? Or that the recording is not appropriate for those sorters? Or something else?
The recording object has these properties:

CommonReferenceRecording: 378 channels - 30.0kHz - 1 segments - 54,261,812 samples - 1,808.73s (30.15 minutes) - int16 dtype - 38.20 GiB

[alejoe91]

Hi @vncntprvst

combinato is only for single-channel electrodes, so it's expected that it fails.

I recently updated herdingspikes, since there was a missing dependency. I think that since you're not specifying the label, it might be using a cached version. Can you try to rerun with docker_image="spikeinterface/herdingspikes-base:0.4.6"?

[vncntprvst]

Hi @alejoe91
I forgot about combinato, thanks for the reminder.

I ran herdingspikes with docker_image="spikeinterface/herdingspikes-base:0.4.6" but it failed again:

print(rec)
BinaryFolderRecording: 378 channels - 30.0kHz - 1 segments - 54,261,812 samples 
                       1,808.73s (30.15 minutes) - int16 dtype - 38.20 GiB

hs_output = sorting_dir / 'sorting_herdingspikes'
container_image="spikeinterface/herdingspikes-base:0.4.6"
params_hs = {}  # use SI defaults

sorting_hs = si.run_sorter(
    sorter_name='herdingspikes',
    recording=rec,
    folder=hs_output,
    remove_existing_folder=True,
    docker_image=container_image, #True, 
    verbose=True,
    **params_hs
)
print('HerdingSpikes units:', len(sorting_hs.get_unit_ids()))

installation_mode='auto' switching to installation_mode: 'github'
Docker: pulling image spikeinterface/herdingspikes-base:0.4.6
Starting container
Installing spikeinterface with github in container
Running herdingspikes sorter inside spikeinterface/herdingspikes-base:0.4.6
Stopping container

---------------------------------------------------------------------------
SpikeSortingError                         Traceback (most recent call last)
Cell In[11], line 6
      3 container_image="spikeinterface/herdingspikes-base:0.4.6"
      4 params_hs = {}  # use SI defaults
----> 6 sorting_hs = si.run_sorter(
      7     sorter_name='herdingspikes',
      8     recording=rec,
      9     folder=hs_output,
     10     remove_existing_folder=True,
     11     docker_image=container_image, #True, 
     12     verbose=True,
     13     **params_hs
     14 )
     15 print('HerdingSpikes units:', len(sorting_hs.get_unit_ids()))

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:204, in run_sorter(sorter_name, recording, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, docker_image, singularity_image, delete_container_files, with_output, **sorter_params)
    198         if not has_spython():
    199             raise RuntimeError(
    200                 "The python `spython` package must be installed to "
    201                 "run singularity. Install with `pip install spython`"
    202             )
--> 204     return run_sorter_container(
    205         container_image=container_image,
    206         mode=mode,
    207         **common_kwargs,
    208     )
    210 return run_sorter_local(**common_kwargs)

File ~/mambaforge/envs/si_env/lib/python3.10/site-packages/spikeinterface/sorters/runsorter.py:681, in run_sorter_container(sorter_name, recording, mode, container_image, folder, remove_existing_folder, delete_output_folder, verbose, raise_error, with_output, delete_container_files, extra_requirements, installation_mode, spikeinterface_version, spikeinterface_folder_source, **sorter_params)
    679 if run_error:
    680     if raise_error:
--> 681         raise SpikeSortingError(f"Spike sorting in {mode} failed with the following error:\n{run_sorter_output}")
    682 else:
    683     if with_output:

SpikeSortingError: Spike sorting in docker failed with the following error: