Change some default parameters for better user experience

src/spikeinterface/core/sparsity.py

@@ -102,7 +102,11 @@ def __init__(self, mask, unit_ids, channel_ids):
 
         self.num_channels = self.channel_ids.size
         self.num_units = self.unit_ids.size
-        self.max_num_active_channels = self.mask.sum(axis=1).max()
+        if self.mask.shape[0]:
+            self.max_num_active_channels = self.mask.sum(axis=1).max()
+        else:
+            # empty sorting without units
+            self.max_num_active_channels = 0
 
     def __repr__(self):
         density = np.mean(self.mask)

src/spikeinterface/core/tests/test_waveform_extractor.py

@@ -556,4 +556,5 @@ def test_non_json_object():
     # test_portability()
     # test_recordingless()
     # test_compute_sparsity()
-    test_non_json_object()
+    # test_non_json_object()
+    test_empty_sorting()

src/spikeinterface/core/waveform_extractor.py

@@ -1457,13 +1457,13 @@ def extract_waveforms(
     folder=None,
     mode="folder",
     precompute_template=("average",),
-    ms_before=3.0,
-    ms_after=4.0,
+    ms_before=1.0,
+    ms_after=2.0,
     max_spikes_per_unit=500,
     overwrite=False,
     return_scaled=True,
     dtype=None,
-    sparse=False,
+    sparse=True,
     sparsity=None,
     num_spikes_for_sparsity=100,
     allow_unfiltered=False,
@@ -1507,7 +1507,7 @@ def extract_waveforms(
         If True and recording has gain_to_uV/offset_to_uV properties, waveforms are converted to uV.
     dtype: dtype or None
         Dtype of the output waveforms. If None, the recording dtype is maintained.
-    sparse: bool (default False)
+    sparse: bool, default: True
         If True, before extracting all waveforms the `precompute_sparsity()` function is run using
         a few spikes to get an estimate of dense templates to create a ChannelSparsity object.
         Then, the waveforms will be sparse at extraction time, which saves a lot of memory.
@@ -1726,6 +1726,7 @@ def precompute_sparsity(
             max_spikes_per_unit=num_spikes_for_sparsity,
             return_scaled=False,
             allow_unfiltered=allow_unfiltered,
+            sparse=False,
             **job_kwargs,
         )
         local_sparsity = compute_sparsity(local_we, **sparse_kwargs)

src/spikeinterface/postprocessing/correlograms.py

@@ -137,8 +137,8 @@ def compute_crosscorrelogram_from_spiketrain(spike_times1, spike_times2, window_
 def compute_correlograms(
     waveform_or_sorting_extractor,
     load_if_exists=False,
-    window_ms: float = 100.0,
-    bin_ms: float = 5.0,
+    window_ms: float = 50.0,
+    bin_ms: float = 1.0,
     method: str = "auto",
 ):
     """Compute auto and cross correlograms.

src/spikeinterface/postprocessing/unit_localization.py

@@ -96,7 +96,7 @@ def get_extension_function():
 
 
 def compute_unit_locations(
-    waveform_extractor, load_if_exists=False, method="center_of_mass", outputs="numpy", **method_kwargs
+    waveform_extractor, load_if_exists=False, method="monopolar_triangulation", outputs="numpy", **method_kwargs
 ):
     """
     Localize units in 2D or 3D with several methods given the template.

src/spikeinterface/sorters/runsorter.py

@@ -91,7 +91,7 @@ def run_sorter(
     sorter_name: str,
     recording: BaseRecording,
     output_folder: Optional[str] = None,
-    remove_existing_folder: bool = True,
+    remove_existing_folder: bool = False,
     delete_output_folder: bool = False,
     verbose: bool = False,
     raise_error: bool = True,

src/spikeinterface/widgets/_legacy_mpl_widgets/collisioncomp.py

@@ -43,6 +43,8 @@ def plot(self):
         self._do_plot()
 
     def _do_plot(self):
+        from matplotlib import pyplot as plt
+
         fig = self.figure
 
         for ax in fig.axes:
@@ -177,6 +179,8 @@ def plot(self):
 
     def _do_plot(self):
         import sklearn
+        import matplotlib.pyplot as plt
+        import matplotlib
 
         # compute similarity
         # take index of template (respect unit_ids order)