@@ 33,10 33,10 @@ def test_supervision_json(client):
     res = client.get('/series/supervision?name=test-edited')
     df = pd.read_json(res.text, orient='index')
     assert_df("""
-                           series  markers
-2020-01-01 00:00:00+00:00       0    False
-2020-01-02 00:00:00+00:00       1    False
-2020-01-03 00:00:00+00:00      42     True
+                           markers  series
+2020-01-01 00:00:00+00:00    False       0
+2020-01-02 00:00:00+00:00    False       1
+2020-01-03 00:00:00+00:00     True      42
 """, df)
 
     res = client.get('/series/supervision?name=test-edited&format=tshpack')

          
@@ 86,7 86,7 @@ 2020-01-03 00:00:00+00:00     True
 """, markers)
 
 
-def test_get_by_horizon(client):
+def test_edited_by_horizon(client):
     ts = genserie(utcdt(2023, 1, 1), 'D', 60)
     client.patch('/series/state', params={
         'name': 'horizon',

          
@@ 107,10 107,10 @@ def test_get_by_horizon(client):
         )
     })
     assert res.json == {
-        '2023-01-30T00:00:00.000Z': {'markers': False, 'series': 29.0},
-        '2023-01-31T00:00:00.000Z': {'markers': False, 'series': 30.0},
-        '2023-02-01T00:00:00.000Z': {'markers': False, 'series': 31.0},
-        '2023-02-02T00:00:00.000Z': {'markers': False, 'series': 32.0}
+        '2023-01-30T00:00:00+00:00': {'markers': False, 'series': 29.0},
+        '2023-01-31T00:00:00+00:00': {'markers': False, 'series': 30.0},
+        '2023-02-01T00:00:00+00:00': {'markers': False, 'series': 31.0},
+        '2023-02-02T00:00:00+00:00': {'markers': False, 'series': 32.0}
     }
     df = pd.read_json(res.text, orient='index')
     ts = df['series']

          
@@ 150,3 150,26 @@ 2023-01-31 00:00:00+00:00    False
 2023-02-01 00:00:00+00:00    False
 2023-02-02 00:00:00+00:00     True
 """, marker)
+
+
+def test_edited_with_timezone(client):
+    ts = genserie(utcdt(2023, 1, 1), 'D', 5)
+    client.patch('/series/state', params={
+        'name': 'withtz',
+        'series': util.tojson(ts),
+        'author': 'Babar',
+        'insertion_date': utcdt(2023, 1, 1),
+        'tzaware': util.tzaware_series(ts),
+        'supervision': json.dumps(False),
+    })
+    res = client.get('/series/supervision', params={
+        'name': 'withtz',
+        'tzone': 'Europe/Paris'
+    })
+    assert res.json == {
+        '2023-01-01T01:00:00+01:00': {'markers': False, 'series': 0.0},
+        '2023-01-02T01:00:00+01:00': {'markers': False, 'series': 1.0},
+        '2023-01-03T01:00:00+01:00': {'markers': False, 'series': 2.0},
+        '2023-01-04T01:00:00+01:00': {'markers': False, 'series': 3.0},
+        '2023-01-05T01:00:00+01:00': {'markers': False, 'series': 4.0}
+    }

          

@@ 58,6 58,10 @@ edited.add_argument(
     help='re-index series on a inferred frequency'
 )
 edited.add_argument(
+    'tzone', type=str, default='UTC',
+    help='Convert tz-aware series into this time zone before sending'
+)
+edited.add_argument(
     '_keep_nans', type=inputs.boolean, default=False,
     help='keep erasure information'
 )

          
@@ 107,18 111,21 @@ class supervision_httpapi(httpapi):
                     inferred_freq=args.get('inferred_freq'),
                     _keep_nans=args._keep_nans
                 )
-                metadata = tsa.metadata(args.name, all=True)
-                assert metadata is not None, f'series {args.name} has no metadata'
+                metadata = tsa.internal_metadata(args.name)
+                if metadata['tzaware'] and args.tzone.upper() != 'UTC':
+                    series.index = series.index.tz_convert(args.tzone)
+                    markers.index = markers.index.tz_convert(args.tzone)
 
                 if args.format == 'json':
                     if series is not None:
                         df = pd.DataFrame()
                         df['series'] = series
                         df['markers'] = markers
-                        response = make_response(
-                            df.to_json(orient='index',
-                                       date_format='iso')
-                        )
+                        out = {
+                            k.isoformat(): v
+                            for k, v in df.to_dict(orient='index').items()
+                        }
+                        response = make_response(out)
                     else:
                         response = make_response('null')
                     response.headers['Content-Type'] = 'text/json'

          

@@ 354,7 354,8 @@ class timeseries(basets):
 
         mask_manual = pd.Series(
             [False] * len(unionindex),
-            index=unionindex
+            index=unionindex,
+            dtype='object'
         )
         if manual is not None:
             mask_manual[manual.index] = True