np.NaN -> np.nan for numpy 2.0 support

ibllib/io/extractors/fibrephotometry.py

@@ -225,7 +225,7 @@ def _extract(self, light_source_map=None, collection=None, regions=None, **kwarg
         regions = regions or [k for k in fp_data['raw'].keys() if 'Region' in k]
         out_df = fp_data['raw'].filter(items=regions, axis=1).sort_index(axis=1)
         out_df['times'] = ts
-        out_df['wavelength'] = np.NaN
+        out_df['wavelength'] = np.nan
         out_df['name'] = ''
         out_df['color'] = ''
         # Extract channel index

ibllib/io/extractors/opto_trials.py

@@ -16,8 +16,8 @@ class LaserBool(BaseBpodTrialsExtractor):
     def _extract(self, **kwargs):
         _logger.info('Extracting laser datasets')
         # reference pybpod implementation
-        lstim = np.array([float(t.get('laser_stimulation', np.NaN)) for t in self.bpod_trials])
-        lprob = np.array([float(t.get('laser_probability', np.NaN)) for t in self.bpod_trials])
+        lstim = np.array([float(t.get('laser_stimulation', np.nan)) for t in self.bpod_trials])
+        lprob = np.array([float(t.get('laser_probability', np.nan)) for t in self.bpod_trials])
 
         # Karolina's choice world legacy implementation - from Slack message:
         # it is possible that some versions I have used:
@@ -30,9 +30,9 @@ def _extract(self, **kwargs):
         # laserOFF_trials=(optoOUT ==0);
         if 'PROBABILITY_OPTO' in self.settings.keys() and np.all(np.isnan(lstim)):
             lprob = np.zeros_like(lprob) + self.settings['PROBABILITY_OPTO']
-            lstim = np.array([float(t.get('opto_ON_time', np.NaN)) for t in self.bpod_trials])
+            lstim = np.array([float(t.get('opto_ON_time', np.nan)) for t in self.bpod_trials])
             if np.all(np.isnan(lstim)):
-                lstim = np.array([float(t.get('optoOUT', np.NaN)) for t in self.bpod_trials])
+                lstim = np.array([float(t.get('optoOUT', np.nan)) for t in self.bpod_trials])
                 lstim[lstim == 255] = 1
             else:
                 lstim[~np.isnan(lstim)] = 1

ibllib/io/extractors/training_trials.py

@@ -32,7 +32,7 @@ def _extract(self):
         feedbackType = np.zeros(len(self.bpod_trials), np.int64)
         for i, t in enumerate(self.bpod_trials):
             state_names = ['correct', 'error', 'no_go', 'omit_correct', 'omit_error', 'omit_no_go']
-            outcome = {sn: ~np.isnan(t['behavior_data']['States timestamps'].get(sn, [[np.NaN]])[0][0]) for sn in state_names}
+            outcome = {sn: ~np.isnan(t['behavior_data']['States timestamps'].get(sn, [[np.nan]])[0][0]) for sn in state_names}
             assert np.sum(list(outcome.values())) == 1
             outcome = next(k for k in outcome if outcome[k])
             if outcome == 'correct':

ibllib/qc/base.py

@@ -85,7 +85,7 @@ def overall_outcome(outcomes: iter, agg=max) -> spec.QC:
         one.alf.spec.QC
             The overall outcome.
         """
-        outcomes = filter(lambda x: x not in (None, np.NaN), outcomes)
+        outcomes = filter(lambda x: x not in (None, np.nan), outcomes)
         return agg(map(spec.QC.validate, outcomes))
 
     def _set_eid_or_path(self, session_path_or_eid):

ibllib/qc/task_metrics.py

@@ -512,7 +512,7 @@ def compute(self, **kwargs):
         iti = (np.roll(data['stimOn_times'], -1) - data['stimOff_times'])[:-1]
         metric = np.r_[np.nan_to_num(iti, nan=np.inf), np.nan] - 1.
         passed[check] = np.abs(metric) <= 0.1
-        passed[check][-1] = np.NaN
+        passed[check][-1] = np.nan
         metrics[check] = metric
 
         # Checks common to training QC