application of interpolation on dataset

forecaster-cnn/main.py

@@ -322,11 +322,7 @@ class Train():
         ###########
         _start = time.time()
         data = data.round(decimals=2)
-        data = missing_data_handling(data, rolling_mean=True)
-        missing_data = percent_missing(data)
-        print("---Missing resume---")
-        print(missing_data)
-        print("---End resume---")
+        data = missing_data_handling(data, slinear_interpolation=True, drop_all_nan=True)
 
         data = datetime_conversion(data, self.time_column_name)
         data = important_data(data, self.features)

forecaster-cnn/pre_processing/preprocessing.py

@@ -33,7 +33,7 @@ def percent_missing(data):
 
 #TODO here has to be placed a function for handling missing data
 def missing_data_handling(data ,drop_all_nan = False, fill_with_mean = False,
-                          fill_with_median = False, rolling_mean = False, rolling_median = False):
+                          fill_with_median = False, rolling_mean = False, rolling_median = False, slinear_interpolation=False):
 
     def drop_all_nan(data):
         data = data.dropna()
@@ -103,10 +103,7 @@ def missing_data_handling(data ,drop_all_nan = False, fill_with_mean = False,
         data.assign(InterpolateSpline4=data.target.interpolate(method='spline', order=5))
         return data
 
-
-    if drop_all_nan == True:
-        data = drop_all_nan(data)
-    elif fill_with_mean == True:
+    if fill_with_mean == True:
         data = fill_with_mean(data)
     elif fill_with_median == True:
         data = fill_with_median(data)
@@ -136,6 +133,8 @@ def missing_data_handling(data ,drop_all_nan = False, fill_with_mean = False,
         data = spline_interpolate4(data)
     elif spline_interpolate5 == True:
         data = spline_interpolate5(data)
+    elif drop_all_nan == True:
+        data = drop_all_nan(data)
     else:
         pass