fix apply_estimator for num_peaks feature

py_neuromodulation/features/sharpwaves.py

@@ -288,14 +288,23 @@ def calc_feature(self, data: np.ndarray) -> dict:
 
             # the key_name stays, since the estimator function stays between peaks and troughs
             for key_name, estimator in self.estimator_key_map.items():
-                if "num_peaks" in key_name:
+                if len(dict_ch_features[key_name]) == 0:
+                    # might happen if num_peaks was written in estimator
+                    # e.g. estimator["mean"] = ["num_peaks"]
+                    # for conveniance this doesn't raise an exception
                     continue
+
                 feature_results[key_name] = estimator(
                     [
                         list(dict_ch_features[key_name].values())[0],
                         list(dict_ch_features[key_name].values())[1],
                     ]
                 )
+            # add here also the num_peaks features
+            if self.sw_settings.sharpwave_features.num_peaks:
+                key_name = f"{ch_name}_Sharpwave_num_peaks_{filter_name}"
+                feature_results[key_name] = np_mean([dict_ch_features[key_name]["Peak"],
+                                                    dict_ch_features[key_name]["Trough"]])
         else:
             # otherwise, save all write all "flattened" key value pairs in feature_results
             for key, subdict in dict_ch_features.items():