search

SpikeInterface / spikeinterface

A Python-based module for creating flexible and robust spike sorting pipelines.
https://spikeinterface.readthedocs.io
MIT License
527 stars 187 forks source link

Scale setting for sorting analyzer extension #2751

Closed Hobart10 closed 6 months ago

Hobart10 commented 6 months ago

Hi, thank you for developing and maintaining this wonderful module and community! I had several issues in computing extension spike_amplitudes and exporting to phy with the extension. Phy exporting requires sorting analyzer parameter return_scaled to be False, but seems the default parameter for spike_amplitudes computation is True and there are no option to modify. And if I don't specify the dtype of waveforms extension to 'int16', the scaling error will happen there. Spikeinterface version: lastest; kilosort 4 version: 4.0.4. Any help of how I can export will be highly appreciated!

Error info: ValueError: get_dense_templates_array: return_scaled=True is not possible SortingAnalyzer has the reverse

Code snippet:

    sorting_KS4 = si.run_sorter(sorter_name='kilosort4', recording=rec_saved, 
                                output_folder=Path(folder_out) / 'results_KS4', remove_existing_folder=True,
                                verbose=False, **ks4_params_modSet_NX)

    folder_analyzer = Path(folder_out) / "sortingAnalyzer_KS4";
    sorting_analyzer = si.create_sorting_analyzer(sorting=sorting_KS4, recording=rec_saved, 
                                          folder=folder_analyzer, format="binary_folder", sparse=False, 
                                          return_scaled=False, overwrite=True, **global_job_kwargs)
    sorting_analyzer.compute("random_spikes")
    sorting_analyzer.compute("waveforms", dtype='int16')
    sorting_analyzer.compute("templates", operators=["average", "median", "std"])
    sorting_analyzer.compute("spike_amplitudes", peak_sign="both")
    sorting_analyzer.compute("template_similarity")

    si.export_to_phy(sorting_analyzer, output_folder=folder_out / 'phyOut_KS4', 
                 compute_pc_features=True, compute_amplitudes=True, 
                 copy_binary=False, remove_if_exists=True, peak_sign = "both", **global_job_kwargs)
Hobart10 commented 6 months ago

Trace back of the error:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
File [d:\YK_pyfile\SI\SI_Proc\SI_procFunc_KS4.py:182](file:///D:/YK_pyfile/SI/SI_Proc/SI_procFunc_KS4.py:182), in SI_sort(folder_out, rec_saved)
--> [182](file:///D:/YK_pyfile/SI/SI_Proc/SI_procFunc_KS4.py:182) sorting_analyzer.compute("spike_amplitudes", peak_sign="both")
    [183](file:///D:/YK_pyfile/SI/SI_Proc/SI_procFunc_KS4.py:183) sorting_analyzer.compute("template_similarity")

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\sortinganalyzer.py:853](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:853), in SortingAnalyzer.compute(self, input, save, extension_params, **kwargs)
    [806](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:806) """
    [807](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:807) Compute one extension or several extensiosn.
    [808](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:808) Internally calls compute_one_extension() or compute_several_extensions() depending on the input type.
   (...)
    [850](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:850) )
    [851](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:851) """
    [852](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:852) if isinstance(input, str):
--> [853](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:853)     return self.compute_one_extension(extension_name=input, save=save, **kwargs)
    [854](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:854) elif isinstance(input, dict):
    [855](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:855)     params_, job_kwargs = split_job_kwargs(kwargs)

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\sortinganalyzer.py:931](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:931), in SortingAnalyzer.compute_one_extension(self, extension_name, save, **kwargs)
    [929](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:929) extension_instance = extension_class(self)
    [930](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:930) extension_instance.set_params(save=save, **params)
--> [931](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:931) extension_instance.run(save=save, **job_kwargs)
    [933](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:933) self.extensions[extension_name] = extension_instance
    [935](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:935) return extension_instance

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\sortinganalyzer.py:1532](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1532), in AnalyzerExtension.run(self, save, **kwargs)
   [1529](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1529)     self._save_params()
   [1530](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1530)     self._save_importing_provenance()
-> [1532](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1532) self._run(**kwargs)
   [1534](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1534) if save and not self.sorting_analyzer.is_read_only():
   [1535](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1535)     self._save_data(**kwargs)

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\postprocessing\spike_amplitudes.py:112](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:112), in ComputeSpikeAmplitudes._run(self, **job_kwargs)
    [110](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:110) def _run(self, **job_kwargs):
    [111](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:111)     job_kwargs = fix_job_kwargs(job_kwargs)
--> [112](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:112)     nodes = self.get_pipeline_nodes()
    [113](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:113)     amps = run_node_pipeline(
    [114](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:114)         self.sorting_analyzer.recording,
    [115](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:115)         nodes,
   (...)
    [118](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:118)         gather_mode="memory",
    [119](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:119)     )
    [120](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:120)     self.data["amplitudes"] = amps

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\sortinganalyzer.py:1700](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1700), in AnalyzerExtension.get_pipeline_nodes(self)
   [1696](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1696) def get_pipeline_nodes(self):
   [1697](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1697)     assert (
   [1698](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1698)         self.use_nodepipeline
   [1699](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1699)     ), "AnalyzerExtension.get_pipeline_nodes() must be called only when use_nodepipeline=True"
-> [1700](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/sortinganalyzer.py:1700)     return self._get_pipeline_nodes()

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\postprocessing\spike_amplitudes.py:92](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:92), in ComputeSpikeAmplitudes._get_pipeline_nodes(self)
     [89](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:89) peak_sign = self.params["peak_sign"]
     [90](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:90) return_scaled = self.sorting_analyzer.return_scaled
---> [92](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:92) extremum_channels_indices = get_template_extremum_channel(
     [93](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:93)     self.sorting_analyzer, peak_sign=peak_sign, outputs="index"
     [94](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:94) )
     [95](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:95) peak_shifts = get_template_extremum_channel_peak_shift(self.sorting_analyzer, peak_sign=peak_sign)
     [97](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:97) spike_retriever_node = SpikeRetriever(
     [98](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:98)     recording, sorting, channel_from_template=True, extremum_channel_inds=extremum_channels_indices
     [99](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:99) )

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\template_tools.py:150](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:150), in get_template_extremum_channel(templates_or_sorting_analyzer, peak_sign, mode, outputs)
    [147](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:147) unit_ids = templates_or_sorting_analyzer.unit_ids
    [148](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:148) channel_ids = templates_or_sorting_analyzer.channel_ids
--> [150](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:150) peak_values = get_template_amplitudes(templates_or_sorting_analyzer, peak_sign=peak_sign, mode=mode)
    [151](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:151) extremum_channels_id = {}
    [152](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:152) extremum_channels_index = {}

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\template_tools.py:90](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:90), in get_template_amplitudes(templates_or_sorting_analyzer, peak_sign, mode, return_scaled)
     [86](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:86) before = _get_nbefore(templates_or_sorting_analyzer)
     [88](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:88) peak_values = {}
---> [90](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:90) templates_array = get_dense_templates_array(templates_or_sorting_analyzer, return_scaled=return_scaled)
     [92](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:92) for unit_ind, unit_id in enumerate(unit_ids):
     [93](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:93)     template = templates_array[unit_ind, :, :]

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\template_tools.py:30](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:30), in get_dense_templates_array(one_object, return_scaled)
     [28](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:28) elif isinstance(one_object, SortingAnalyzer):
     [29](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:29)     if return_scaled != one_object.return_scaled:
---> [30](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:30)         raise ValueError(
     [31](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:31)             f"get_dense_templates_array: return_scaled={return_scaled} is not possible SortingAnalyzer has the reverse"
     [32](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:32)         )
     [33](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:33)     ext = one_object.get_extension("templates")
     [34](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:34)     if ext is not None:

ValueError: get_dense_templates_array: return_scaled=True is not possible SortingAnalyzer has the reverse
zm711 commented 6 months ago

@Hobart10,

You're right. We recently switched to return_scaled as a variable at the sorting_analyzer level. We missed switching over the init here

https://github.com/SpikeInterface/spikeinterface/blob/163ab282645e62fbfd75cd845eb7755da187c787/src/spikeinterface/postprocessing/spike_amplitudes.py#L146-L157

We will fix it!

Hobart10 commented 6 months ago

Thank you!!

zm711 commented 6 months ago

Actually I think I tracked down the real issue. It is that we use the default return_scaled=True here

https://github.com/SpikeInterface/spikeinterface/blob/163ab282645e62fbfd75cd845eb7755da187c787/src/spikeinterface/core/template_tools.py#L56-L61

Because here we don't give the explicit return_scaled.

https://github.com/SpikeInterface/spikeinterface/blob/163ab282645e62fbfd75cd845eb7755da187c787/src/spikeinterface/core/template_tools.py#L150

I will submit a PR in little bit :)

zm711 commented 6 months ago

@alejoe91 @samuelgarcia

So we need to be able to use return_scaled in get_template_amplitudes but the function needs to accept SortingAnalyzer or Templates, but Templates doesn't have a global return_scaled like SortingAnalyzer. How do we want to deal with this?

alejoe91 commented 6 months ago

Maybe just ignore the return_scaled if a Templates object is passed?

zm711 commented 6 months ago

So in the PR I just sent return_scaled to True (the previous behavior) if a Templates object is passed since it is a required argument for the function. Let me know if you want something else.

@Hobart10, if you want to test this for us you can install from the linked PR and see if that fixes your issue. We might be missing another spot so it would be cool to have someone test this so we can make sure all steps are working :)

Hobart10 commented 6 months ago

Hi Zach @zm711, thank you for the fixing! I tested with the PR and got same error and I think it's because return_scaled didn't pass to get_dense_templates_array when calling from get_template_extremum_channel_peak_shift, and default return_scaled in get_dense_templates_array is True. https://github.com/SpikeInterface/spikeinterface/blob/e440d8a6dab688ad8a2ffab1e82b3ce530626a15/src/spikeinterface/core/template_tools.py#L199

Therefore this error raises: https://github.com/SpikeInterface/spikeinterface/blob/e440d8a6dab688ad8a2ffab1e82b3ce530626a15/src/spikeinterface/core/template_tools.py#L26-L32

Trace back of the error:

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\postprocessing\spike_amplitudes.py:95](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:95), in ComputeSpikeAmplitudes._get_pipeline_nodes(self)
     [90](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:90) return_scaled = self.sorting_analyzer.return_scaled
     [92](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:92) extremum_channels_indices = get_template_extremum_channel(
     [93](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:93)     self.sorting_analyzer, peak_sign=peak_sign, outputs="index"
     [94](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:94) )
---> [95](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:95) peak_shifts = get_template_extremum_channel_peak_shift(self.sorting_analyzer, peak_sign=peak_sign)
     [97](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:97) spike_retriever_node = SpikeRetriever(
     [98](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:98)     recording, sorting, channel_from_template=True, extremum_channel_inds=extremum_channels_indices
     [99](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:99) )
    [100](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:100) spike_amplitudes_node = SpikeAmplitudeNode(
    [101](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:101)     recording,
    [102](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:102)     parents=[spike_retriever_node],
   (...)
    [105](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:105)     return_scaled=return_scaled,
    [106](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/postprocessing/spike_amplitudes.py:106) )

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\template_tools.py:199](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:199), in get_template_extremum_channel_peak_shift(templates_or_sorting_analyzer, peak_sign)
    [195](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:195) extremum_channels_ids = get_template_extremum_channel(templates_or_sorting_analyzer, peak_sign=peak_sign)
    [197](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:197) shifts = {}
--> [199](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:199) templates_array = get_dense_templates_array(templates_or_sorting_analyzer)
    [201](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:201) for unit_ind, unit_id in enumerate(unit_ids):
    [202](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:202)     template = templates_array[unit_ind, :, :]

File [~\.conda\envs\SI\spikeinterface\src\spikeinterface\core\template_tools.py:30](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:30), in get_dense_templates_array(one_object, return_scaled)
     [28](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:28) elif isinstance(one_object, SortingAnalyzer):
     [29](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:29)     if return_scaled != one_object.return_scaled:
---> [30](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:30)         raise ValueError(
     [31](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:31)             f"get_dense_templates_array: return_scaled={return_scaled} is not possible SortingAnalyzer has the reverse"
     [32](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:32)         )
     [33](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:33)     ext = one_object.get_extension("templates")
     [34](https://file+.vscode-resource.vscode-cdn.net/d%3A/YK_pyfile/SI/SI_Proc/~/.conda/envs/SI/spikeinterface/src/spikeinterface/core/template_tools.py:34)     if ext is not None:

ValueError: get_dense_templates_array: return_scaled=True is not possible SortingAnalyzer has the reverse
zm711 commented 6 months ago

Good catch @Hobart10 ! I just patched that and found one other spot in template tools where we needed to grab the return_scaled. Could you try now with the same PR!

Hobart10 commented 6 months ago

@zm711 Thank you!! Now it works in my test!

zm711 commented 6 months ago

@Hobart10 We'll close this one when the PR is merged :)