Move data cropping to the models #335 (#336)

analytics/analytics/analytic_unit_manager.py

@@ -33,9 +33,6 @@ def prepare_data(data: list):
     data = pd.DataFrame(data, columns=['timestamp', 'value'])
     data['timestamp'] = pd.to_datetime(data['timestamp'], unit='ms')
     data.fillna(value = np.nan, inplace = True)
-    if not np.isnan(data['value'].min()):
-        data['value'] = data['value'] - min(data['value'])
-
     return data
 
 

analytics/analytics/models/drop_model.py

@@ -28,7 +28,8 @@ class DropModel(Model):
         }
 
     def do_fit(self, dataframe: pd.DataFrame, segments: list) -> None:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         confidences = []
         convolve_list = []
         drop_height_list = []
@@ -101,7 +102,8 @@ class DropModel(Model):
             self.state['conv_del_max'] = self.state['WINDOW_SIZE']
 
     def do_detect(self, dataframe: pd.DataFrame) -> list:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         possible_drops = utils.find_drop(data, self.state['DROP_HEIGHT'], self.state['DROP_LENGTH'] + 1)
 
         return self.__filter_detection(possible_drops, data)

analytics/analytics/models/general_model.py

@@ -28,7 +28,8 @@ class GeneralModel(Model):
         self.all_conv = []
 
     def do_fit(self, dataframe: pd.DataFrame, segments: list) -> None:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         convolve_list = []
         patterns_list = []
         for segment in segments:
@@ -79,7 +80,8 @@ class GeneralModel(Model):
             self.state['conv_del_max'] = self.state['WINDOW_SIZE']
 
     def do_detect(self, dataframe: pd.DataFrame) -> list:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         pat_data = self.model_gen
         y = max(pat_data)
 

analytics/analytics/models/jump_model.py

@@ -29,7 +29,8 @@ class JumpModel(Model):
         }
 
     def do_fit(self, dataframe: pd.DataFrame, segments: list) -> None:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         confidences = []
         convolve_list = []
         jump_height_list = []
@@ -102,7 +103,8 @@ class JumpModel(Model):
             self.state['conv_del_max'] = self.state['WINDOW_SIZE']
 
     def do_detect(self, dataframe: pd.DataFrame) -> list:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         possible_jumps = utils.find_jump(data, self.state['JUMP_HEIGHT'], self.state['JUMP_LENGTH'] + 1)
 
         return self.__filter_detection(possible_jumps, data)

analytics/analytics/models/peak_model.py

@@ -28,7 +28,8 @@ class PeakModel(Model):
         }
 
     def do_fit(self, dataframe: pd.DataFrame, segments: list) -> None:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         confidences = []
         convolve_list = []
         patterns_list = []
@@ -87,7 +88,8 @@ class PeakModel(Model):
             self.state['conv_del_max'] = self.state['WINDOW_SIZE']
 
     def do_detect(self, dataframe: pd.DataFrame):
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         window_size = int(len(data)/SMOOTHING_COEFF) #test ws on flat data
         all_maxs = argrelextrema(np.array(data), np.greater)[0]
 

analytics/analytics/models/trough_model.py

@@ -28,7 +28,8 @@ class TroughModel(Model):
         }
 
     def do_fit(self, dataframe: pd.DataFrame, segments: list) -> None:
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         confidences = []
         convolve_list = []
         patterns_list = []
@@ -88,7 +89,8 @@ class TroughModel(Model):
             self.state['conv_del_max'] = self.state['WINDOW_SIZE']
 
     def do_detect(self, dataframe: pd.DataFrame):
-        data = dataframe['value']
+        data = utils.cut_dataframe(dataframe)
+        data = data['value']
         window_size = int(len(data)/SMOOTHING_COEFF) #test ws on flat data
         all_mins = argrelextrema(np.array(data), np.less)[0]
 

analytics/analytics/utils/common.py

@@ -273,4 +273,8 @@ def pattern_intersection(segment_data: list, median: float, pattern_type: str) -
 
     return [x for (idx, x) in enumerate(center_index) if idx not in delete_index]
 
-    
+def cut_dataframe(data: pd.DataFrame) -> pd.DataFrame:
+    data_min = data['value'].min()
+    if not np.isnan(data_min) and data_min > 0:
+        data['value'] = data['value'] - data_min
+    return data

analytics/tests/test_dataset.py

@@ -121,10 +121,10 @@ class TestDataset(unittest.TestCase):
         data_none = [[1523889000000, None], [1523889000001, None], [1523889000002, None]]
         return_data_nan = prepare_data(data_nan)
         return_data_none = prepare_data(data_none)
-        for item in return_data_nan:
-            self.assertTrue(np.isnan(item.value))
-        for item in return_data_none:
-            self.assertTrue(np.isnan(item.value))
+        for item in return_data_nan.value:
+            self.assertTrue(np.isnan(item))
+        for item in return_data_none.value:
+            self.assertTrue(np.isnan(item))
 
 if __name__ == '__main__':
     unittest.main()