diff --git a/docs/FunctionDescriptions.md b/docs/FunctionDescriptions.md
index ec99fe89fce8c584b73ad6058c85a8d0c35a5634..5c677ca4d705260559165bfbfb0c63aad3371079 100644
--- a/docs/FunctionDescriptions.md
+++ b/docs/FunctionDescriptions.md
@@ -602,7 +602,7 @@ This Function uses pre-trained machine-learning model objects for flagging.
 This requires training a model by use of the [training script](../ressources/machine_learning/train_machine_learning.py) provided. 
 For flagging, inputs to the model are the data of the variable of interest, 
 data of reference variables and the automatic flags that were assigned by other 
-tests inside SaQC. 
+tests inside SaQC. Therefore, this function should be defined last in the config-file, i.e. it should be the last test that is executed.
 Internally, context information for each point is gathered in form of moving 
 windows. The size of the moving windows for counting of the surrounding 
 automatic flags and for calculation of gradients in the data is specified by 
diff --git a/ressources/machine_learning/train_machine_learning.py b/ressources/machine_learning/train_machine_learning.py
index a6e3e35a725459639487805810e1e6d226eb5419..17dbc8535f8d9a2232256a2caf8f3fb20499d3ff 100644
--- a/ressources/machine_learning/train_machine_learning.py
+++ b/ressources/machine_learning/train_machine_learning.py
@@ -4,12 +4,14 @@ import random  # for random sampling of training/test
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.metrics import recall_score, precision_score, classification_report
 import joblib  # for saving of model objects
-
+import os
+import time
+import datetime
 ###--------------------
 ### EXAMPLE PARAMETRIZATION:
 ###--------------------
 
-# pd.options.mode.chained_assignment = None  # default='warn'
+#pd.options.mode.chained_assignment = None  # default='warn'
 # data = pd.read_feather("data/sm/02_data.feather")
 # data = data.reset_index()#data.index has to be reset as I use row nos only for indexing
 #
@@ -23,9 +25,8 @@ import joblib  # for saving of model objects
 # references = ["Var1","Var2"]
 # window_values = 20
 # window_flags = 20
-# modelname="testmodel"
-# #groupvar = 0.2
-# path = "saqc/ressources/machine_learning/models/"
+# modelname="name"
+# path = "models/"
 # sensor_field="SensorID"
 # group_field = "GroupVar"
 
@@ -71,6 +72,25 @@ def trainML(
     :param testratio                    A float denoting the ratio of the test- vs. training-set to be drawn from the data, e.g. 0.3
     """
 
+    def _refCalc(reference, window_values):
+        # Helper function for calculation of moving window values
+        outdata = pd.DataFrame()
+        name = reference.name
+        # derive gradients from reference series
+        outdata[name + "_Dt_1"] = reference - reference.shift(1)  # gradient t vs. t-1
+        outdata[name + "_Dt1"] = reference - reference.shift(-1)  # gradient t vs. t+1
+        # moving mean of gradients var1 and var2 before/after
+        outdata[name + "_Dt_" + str(window_values)] = (
+            outdata[name + "_Dt_1"].rolling(window_values, center=False).mean()
+        )  # mean gradient t to t-window
+        outdata[name + "_Dt" + str(window_values)] = (
+            outdata[name + "_Dt_1"]
+            .iloc[::-1]
+            .rolling(window_values, center=False)
+            .mean()[::-1]
+        )  # mean gradient t to t+window
+        return outdata
+
     randomseed = 36
     ### Prepare data, i.e. compute moving windows
     print("Computing time-lags")
@@ -109,7 +129,7 @@ def trainML(
         # Add context information for field+references
         for i in [field] + references:
             sensordf = pd.concat(
-                [sensordf, refCalc(reference=sensordf[i], window_values=window_values)],
+                [sensordf, _refCalc(reference=sensordf[i], window_values=window_values)],
                 axis=1,
             )
 
@@ -128,7 +148,7 @@ def trainML(
     # make column in "traindata" to store predictions
     traindata = traindata.assign(PredMan=0)
     outinfo_df = []
-    resultfile = open(os.path.join(path, modelname + "_resultfile.txt"), "w")
+    resultfile = open(os.path.join(os.getcwd(),path, modelname + "_resultfile.txt"), "w")
     starttime = time.time()
     # For each category of groupvar, fit a separate model
 
@@ -204,7 +224,7 @@ def trainML(
         ] = preds_te
 
     endtime = time.time()
-    print("TIME ELAPSED: " + str(timedelta(seconds=endtime - starttime)) + " min")
+    print("TIME ELAPSED: " + str(datetime.timedelta(seconds=endtime - starttime)) + " hours")
     outinfo_df = pd.DataFrame.from_records(
         outinfo_df,
         columns=[
@@ -230,7 +250,7 @@ def trainML(
     data.PredMan[
         traindata.RowIndex
     ] = traindata.PredMan  # based on RowIndex as NAs were created in traindata
-    data.to_feather("data/sm/03_data_preds")
+    data.to_feather(os.path.join(path, modelname + "_data_preds.feather"))
 
 
 trainML(