diff --git a/scripts/checkPulseData.py b/scripts/checkPulseData.py
index e0d9ad7..d826906 100644
--- a/scripts/checkPulseData.py
+++ b/scripts/checkPulseData.py
@@ -54,7 +54,7 @@
     return np.array(prune(pulse))
 
 
-def monotonic(x):
+def is_monotonic(x):
     if not len(x):
         return False
     dx = np.diff(x)
@@ -93,6 +93,12 @@
 
 
 class Table(dict):
+    '''Dictionary tweaked for conversion into pandas DataFrame
+       
+       Dictionary conversion into DataFrames is faster than 
+       appending rows to a DataFrame, as pointed out here:
+       http://stackoverflow.com/a/17496530/3324012
+    '''
 
     def __init__(self, *keylist):
         super(dict, self).__init__()
@@ -142,20 +148,7 @@
     not_found = StringTemplate('{LAY}/{SEC} on fill {FILL}: calibration step {CS} not found for ns={NS}')
     not_found = not_found.format(LAY=layer)
 
-    # Init data frame
-    data_frame = pd.DataFrame(columns=['Detector',
-                                       'Sector',
-                                       'Fill',
-                                       'N strips',
-                                       'V bias',
-                                       'MPVs'])
-    # dic = {'Detector': [],
-    #        'Sector': [],
-    #        'Fill': [],
-    #        'N strips': [],
-    #        'V bias': [],
-    #        'MPVs': []}
-
+    # Init dictionary for data frame
     table = Table('Detector', 'Sector', 'Fill', 'N strips', 'V bias', 'MPVs')
 
     # Loop on data
@@ -180,19 +173,9 @@
                                                        CS=cstep))
                             pulse.append(None)
                     pulse = build_pulse(pulse)
-                    if monotonic(pulse):
-                        # Append this occurrence to the data frame
-                        # Probably it would be faster to do as suggested here:
-                        # http://stackoverflow.com/a/17496530/3324012
-                        # data_frame.loc[len(data_frame)] = [detector,
-                        #                                    sector,
-                        #                                    fill,
-                        #                                    ns,
-                        #                                    vmap[vstep],
-                        #                                    pulse]
-
-                        table.append(detector, sector, fill, ns, vmap[vstep], pulse)
-
+                    # Append this occurrence to the data frame dictionary
+                    table.append(detector, sector, fill, ns, vmap[vstep], pulse)
+                    if is_monotonic(pulse):
                         if VERBOSE:
                             pulse = [round(v,2) for v in pulse]
                             print(warning.format(NS=ns,
@@ -202,5 +185,10 @@
                                                  PULSE=pulse))
 
     datafile.Close()
-    # data_frame[data_frame['MPVs'].apply(monotonic)]
+    # Create the pandas DataFrame from the dictionary
     data_frame = pd.DataFrame.from_dict(table)
+    # Select the instances where pulse data are monotonic
+    monotonics = data_frame[data_frame['MPVs'].apply(is_monotonic)]
+
+    import pickle
+    pickle.dump(monotonics, open('monotonics.pkl', 'wb'))