diff --git a/PRINS/run_PRINS.py b/PRINS/run_PRINS.py
index 74090a5..b5e7725 100644
--- a/PRINS/run_PRINS.py
+++ b/PRINS/run_PRINS.py
@@ -58,8 +58,9 @@
     m_sys, _, _, _ = instance.run(mint_timeout=args.timeout,
                                   ignore_values=args.ignore_values,
                                   save_pdf=args.save_pdf,
+                                  det = args.det != 'none',
                                   num_workers=args.num_workers,
                                   use_pickle=args.use_pickle)
     if args.det != 'none':
-        PRINS.postprocess(m_sys, determinize_technique=args.det)
+        PRINS.postprocess(m_sys, determinize_technique=args.det, save_pdf=args.save_pdf, output_dir = os.path.join(args.output_dir, args.system))
     logger.info('run_PRINS ends without error(s).')
diff --git a/PRINS/src/main/PRINS.py b/PRINS/src/main/PRINS.py
index 940c218..49d57d5 100644
--- a/PRINS/src/main/PRINS.py
+++ b/PRINS/src/main/PRINS.py
@@ -70,7 +70,7 @@ def __init__(self, system: str, logs, output_dir: str):
             self.l_vectors = convert_df_into_l_vectors(self.logs_df, include_component=True)
             self.components = natsorted(self.logs_df.component.unique())
 
-    def run(self, mint_timeout=3600, mint_param=2, ignore_values=False, save_pdf=True, num_workers=4, use_pickle=False):
+    def run(self, mint_timeout=3600, mint_param=2, ignore_values=False, save_pdf=True, det = False, num_workers=4, use_pickle=False):
         """Run PRINS (main algorithm).
 
         :param mint_timeout: mint timeout (sec) (default=3600)
@@ -169,8 +169,8 @@ def run(self, mint_timeout=3600, mint_param=2, ignore_values=False, save_pdf=Tru
         lc_div_score = len(set(all_components)) / len(all_components)
         print(f'lcDivScore: {lc_div_score:.3f}')
 
-        # (optional) Saving the Results as a pdf file
-        if save_pdf:
+        # (optional) Saving the Results as a pdf file if deterministic technique is not asked
+        if save_pdf and not det:
             if len(m_sys.states) < 1000:
                 m_sys.save_pdf(output_dir=self.output_dir)
             else:
@@ -253,7 +253,7 @@ def partition_log_by_component(l_vector: list) -> list:
         return partitioned_log
 
     @staticmethod
-    def postprocess(m_sys: 'NFA', determinize_technique: str = 'hybrid-1'):
+    def postprocess(m_sys: 'NFA', determinize_technique: str = 'hybrid-1', save_pdf=True, output_dir = "/."):
         """
         Postprocessing function.
         Currently, we only convert NFA to DFA.
@@ -279,4 +279,11 @@ def postprocess(m_sys: 'NFA', determinize_technique: str = 'hybrid-1'):
         print(f'Post-processed model (m_sys): states={len(dfa_sys.states)}, transitions={len(dfa_sys.transitions)}')
         logger.info(f'Post-processed model (m_sys): states={len(dfa_sys.states)}, transitions={len(dfa_sys.transitions)}')
 
+        # (optional) Saving the Results as a pdf file
+        if save_pdf:
+            if len(dfa_sys.states) < 1000:
+                dfa_sys.save_pdf(output_dir)
+            else:
+                print(f'Do not save model.pdf because of too many states: {len(m_sys.states)}')
+
         return dfa_sys, time.time() - start_time