Bug with compute confuision matrix when using 'best' match mode

[HughYau]

When I do the comparision with ground truth data, I found that it often happens error when using 'best' mode, but when I turned it to 'hungarian' mode, it works. It look like some problem when create the pandas dataframe, it should provides 3+13+1 (44) table, but something wrong with its indices so expected (4,5) values

---

ValueError Traceback (most recent call last)
Cell In[19], line 9
7 si.plot_agreement_matrix(comp, ordered=True,backend='matplotlib',figtitle = sorter_name+' Agreement matrix')
8 plt.savefig(output_folder/f'agreement_{sorter_name}.png')
----> 9 si.plot_confusion_matrix(comp,backend='matplotlib',figtitle = sorter_name+' Confusion matrix')
10 plt.savefig(output_folder/f'confusion_{sorter_name}.png')
11 perf = comp.get_performance()

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\widgets\comparison.py:29, in ConfusionMatrixWidget.init(self, gt_comparison, count_text, unit_ticks, backend, **backend_kwargs)
23 def init(self, gt_comparison, count_text=True, unit_ticks=True, backend=None, **backend_kwargs):
24 plot_data = dict(
25 gt_comparison=gt_comparison,
26 count_text=count_text,
27 unit_ticks=unit_ticks,
28 )
---> 29 BaseWidget.init(self, plot_data, backend=backend, **backend_kwargs)

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\widgets\base.py:82, in BaseWidget.init(self, data_plot, backend, immediate_plot, **backend_kwargs)
79 self.backend_kwargs = backend_kwargs_
81 if immediate_plot:
---> 82 self.do_plot()

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\widgets\base.py:103, in BaseWidget.do_plot(self)
101 def do_plot(self):
102 func = getattr(self, f"plot_{self.backend}")
--> 103 func(self.data_plot, **self.backend_kwargs)

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\widgets\comparison.py:41, in ConfusionMatrixWidget.plot_matplotlib(self, data_plot, **backend_kwargs)
37 self.figure, self.axes, self.ax = make_mpl_figure(**backend_kwargs)
39 comp = dp.gt_comparison
---> 41 confusion_matrix = comp.get_confusion_matrix()
42 N1 = confusion_matrix.shape[0] - 1
43 N2 = confusion_matrix.shape[1] - 1

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\comparison\paircomparisons.py:373, in GroundTruthComparison.get_confusion_matrix(self)
364 """
365 Computes the confusion matrix.
366
(...)
370 The confusion matrix
371 """
372 if self._confusion_matrix is None:
--> 373 self._do_confusion_matrix()
374 return self._confusion_matrix

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\comparison\paircomparisons.py:359, in GroundTruthComparison._do_confusion_matrix(self)
356 elif self.match_mode == "best":
357 match_12 = self.best_match_12
--> 359 self._confusion_matrix = do_confusion_matrix(
360 self.event_counts1, self.event_counts2, match_12, self.match_event_count
361 )

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\spikeinterface\comparison\comparisontools.py:720, in do_confusion_matrix(event_counts1, event_counts2, match_12, match_event_count)
716 ordered_units2 = np.hstack([matched_units2, unmatched_units2])
718 import pandas as pd
--> 720 conf_matrix = pd.DataFrame(
721 np.zeros((N1 + 1, N2 + 1), dtype=int),
722 index=list(ordered_units1) + ["FP"],
723 columns=list(ordered_units2) + ["FN"],
724 )
726 for u1 in matched_units1:
727 u2 = match_12[u1]

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\pandas\core\frame.py:816, in DataFrame.init(self, data, index, columns, dtype, copy)
805 mgr = dict_to_mgr(
806 # error: Item "ndarray" of "Union[ndarray, Series, Index]" has no
807 # attribute "name"
(...)
813 copy=_copy,
814 )
815 else:
--> 816 mgr = ndarray_to_mgr(
817 data,
818 index,
819 columns,
820 dtype=dtype,
821 copy=copy,
822 typ=manager,
823 )
825 # For data is list-like, or Iterable (will consume into list)
826 elif is_list_like(data):

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\pandas\core\internals\construction.py:336, in ndarray_to_mgr(values, index, columns, dtype, copy, typ)
331 # _prep_ndarraylike ensures that values.ndim == 2 at this point
332 index, columns = _get_axes(
333 values.shape[0], values.shape[1], index=index, columns=columns
334 )
--> 336 _check_values_indices_shape_match(values, index, columns)
338 if typ == "array":
339 if issubclass(values.dtype.type, str):

File d:\Applications\anaconda3\envs\eeg\lib\site-packages\pandas\core\internals\construction.py:420, in _check_values_indices_shape_match(values, index, columns)
418 passed = values.shape
419 implied = (len(index), len(columns))
--> 420 raise ValueError(f"Shape of passed values is {passed}, indices imply {implied}")

ValueError: Shape of passed values is (4, 4), indices imply (4, 5)

[HughYau]

Also, I found it have relations with 'delta_time' parameter, when I decrease it, it works well