fixing notebook failures

rdtools/soiling.py

@@ -438,7 +438,7 @@ def _calc_result_df(self, trim=False, max_relative_slope_error=500.0, max_negati
             results.loc[filt, "run_slope"] = 0
             results.loc[filt, "run_slope_low"] = 0
             results.loc[filt, "run_slope_high"] = 0
-            # results.loc[filt, "valid"] = False
+            results.loc[filt, "valid"] = False
 
         # Calculate the next inferred start loss from next valid interval
         results["next_inferred_start_loss"] = np.clip(
@@ -465,10 +465,10 @@ def _calc_result_df(self, trim=False, max_relative_slope_error=500.0, max_negati
         results.loc[results.clean_event, "inferred_begin_shift"] = np.clip(
             results.inferred_begin_shift, 0, 1)
         #######################################################################
-
+        '''
         if neg_shift is False:
             results.loc[filt, "valid"] = False
-
+        '''
         if len(results[results.valid]) == 0:
             raise NoValidIntervalError("No valid soiling intervals were found")
         new_start = results.start.iloc[0]
@@ -2952,4 +2952,4 @@ def segmented_soiling_period(
     # Create Series from modelled profile
     sr = pd.Series(z, index=pr.index)
 
-    return sr, cp_date
+    return sr, cp_date