Fix a bunch of deprecations and stylistic problems

Author: twiecki

pyfolio/plotting.py

@@ -645,7 +645,7 @@ def show_perf_stats(returns, factor_returns=None, positions=None,
         perf_stats = pd.DataFrame(perf_stats_all, columns=['Backtest'])
 
     for column in perf_stats.columns:
-        for stat, value in perf_stats[column].iteritems():
+        for stat, value in perf_stats[column].items():
             if stat in STAT_FUNCS_PCT:
                 perf_stats.loc[stat, column] = str(np.round(value * 100,
                                                             1)) + '%'
@@ -1702,7 +1702,7 @@ def cumulate_returns(x):
                 y=monthly_rets.values,
                 color='steelblue')
 
-    locs, labels = plt.xticks()
+    _, labels = plt.xticks()
     plt.setp(labels, rotation=90)
 
     # only show x-labels on year boundary

pyfolio/round_trips.py

@@ -120,7 +120,7 @@ def vwap(transaction):
             transaction.amount.sum()
 
     out = []
-    for sym, t in txn.groupby('symbol'):
+    for _, t in txn.groupby('symbol'):
         t = t.sort_index()
         t.index.name = 'dt'
         t = t.reset_index()
@@ -130,11 +130,11 @@ def vwap(transaction):
             1) != t.order_sign).astype(int).cumsum()
         t['block_time'] = ((t.dt.sub(t.dt.shift(1))) >
                            max_delta).astype(int).cumsum()
-        grouped_price = (t.groupby(('block_dir',
-                                   'block_time'))
+        grouped_price = (t.groupby(['block_dir',
+                                   'block_time'])
                           .apply(vwap))
         grouped_price.name = 'price'
-        grouped_rest = t.groupby(('block_dir', 'block_time')).agg({
+        grouped_rest = t.groupby(['block_dir', 'block_time']).agg({
             'amount': 'sum',
             'symbol': 'first',
             'dt': 'first'})
@@ -265,7 +265,7 @@ def extract_round_trips(transactions,
                                                                  minute=0,
                                                                  second=0))
 
-        tmp = roundtrips.join(pv, on='date', lsuffix='_')
+        tmp = roundtrips.set_index('date').join(pv.set_index('date'), lsuffix='_').reset_index()
 
         roundtrips['returns'] = tmp.pnl / tmp.portfolio_value
         roundtrips = roundtrips.drop('date', axis='columns')
@@ -301,15 +301,17 @@ def add_closing_transactions(positions, transactions):
     # they don't conflict with other round_trips executed at that time.
     end_dt = open_pos.name + pd.Timedelta(seconds=1)
 
-    for sym, ending_val in open_pos.iteritems():
+    for sym, ending_val in open_pos.items():
         txn_sym = transactions[transactions.symbol == sym]
 
         ending_amount = txn_sym.amount.sum()
 
         ending_price = ending_val / ending_amount
-        closing_txn = {'symbol': sym,
-                       'amount': -ending_amount,
-                       'price': ending_price}
+        closing_txn = OrderedDict([
+            ('amount', -ending_amount),
+            ('price', ending_price),
+            ('symbol', sym),
+        ])
 
         closing_txn = pd.DataFrame(closing_txn, index=[end_dt])
         closed_txns = closed_txns.append(closing_txn)

pyfolio/tears.py

@@ -1066,7 +1066,7 @@ def create_capacity_tear_sheet(returns, positions, transactions,
         llt[llt['max_pct_bar_consumed'] > trade_daily_vol_limit * 100])
 
     bt_starting_capital = positions.iloc[0].sum() / (1 + returns.iloc[0])
-    fig, ax_capacity_sweep = plt.subplots(figsize=(14, 6))
+    _, ax_capacity_sweep = plt.subplots(figsize=(14, 6))
     plotting.plot_capacity_sweep(returns, transactions, market_data,
                                  bt_starting_capital,
                                  min_pv=100000,
@@ -1516,7 +1516,7 @@ def create_perf_attrib_tear_sheet(returns,
 
     if factor_partitions is not None:
 
-        for factor_type, partitions in factor_partitions.iteritems():
+        for factor_type, partitions in factor_partitions.items():
 
             columns_to_select = perf_attrib_data.columns.intersection(
                 partitions
@@ -1531,7 +1531,7 @@ def create_perf_attrib_tear_sheet(returns,
             )
             current_section += 1
 
-        for factor_type, partitions in factor_partitions.iteritems():
+        for factor_type, partitions in factor_partitions.items():
 
             perf_attrib.plot_risk_exposures(
                 portfolio_exposures[portfolio_exposures.columns

pyfolio/tests/test_perf_attrib.py

@@ -117,8 +117,8 @@ def test_perf_attrib_simple(self):
 
         expected_perf_attrib_output = pd.DataFrame(
             index=dts,
-            columns=['risk_factor1', 'risk_factor2', 'common_returns',
-                     'specific_returns', 'total_returns'],
+            columns=['risk_factor1', 'risk_factor2', 'total_returns',
+                     'common_returns', 'specific_returns'],
             data={'risk_factor1': [0.025, 0.025],
                   'risk_factor2': [0.025, 0.025],
                   'common_returns': [0.05, 0.05],
@@ -157,8 +157,8 @@ def test_perf_attrib_simple(self):
 
         expected_perf_attrib_output = pd.DataFrame(
             index=dts,
-            columns=['risk_factor1', 'risk_factor2', 'common_returns',
-                     'specific_returns', 'total_returns'],
+            columns=['risk_factor1', 'risk_factor2', 'total_returns',
+                     'common_returns', 'specific_returns'],
             data={'risk_factor1': [0.0, 0.0],
                   'risk_factor2': [0.0, 0.0],
                   'common_returns': [0.0, 0.0],

pyfolio/tests/test_round_trips.py

@@ -152,12 +152,12 @@ def test_add_closing_trades(self):
                                        [-5, 10, 'A'],
                                        [-1, 10, 'B']],
                                  columns=['amount', 'price', 'symbol'],
-                                 index=[dates[:3]])
+                                 index=dates[:3])
         positions = DataFrame(data=[[20, 10, 0],
                                     [-30, 10, 30],
                                     [-60, 0, 30]],
                               columns=['A', 'B', 'cash'],
-                              index=[dates[:3]])
+                              index=dates[:3])
 
         expected_ix = dates[:3].append(DatetimeIndex([dates[2] +
                                                       Timedelta(seconds=1)]))

pyfolio/tests/test_timeseries.py

@@ -60,7 +60,7 @@ def test_gen_drawdown_table_relative(
             first_expected_recovery, first_net_drawdown,
             second_expected_peak, second_expected_valley,
             second_expected_recovery, second_net_drawdown
-            ):
+    ):
 
         rets = px.pct_change()
 
@@ -264,11 +264,12 @@ class TestStats(TestCase):
     dt_2 = pd.date_range('2000-1-3', periods=8, freq='D')
 
     @parameterized.expand([
-        (simple_rets[:5], 2, '[nan, inf, inf, 11.224972160321828, inf]')
+        (simple_rets[:5], 2, [np.nan, np.inf, np.inf, 11.224972160321, np.inf])
     ])
     def test_sharpe_2(self, returns, rolling_sharpe_window, expected):
-        self.assertEqual(str(timeseries.rolling_sharpe(
-            returns, rolling_sharpe_window).values.tolist()), expected)
+        np.testing.assert_array_almost_equal(timeseries.rolling_sharpe(returns,
+                                                                       rolling_sharpe_window).values,
+                                             np.asarray(expected))
 
     @parameterized.expand([
         (simple_rets[:5], simple_benchmark, 2, 0)

pyfolio/timeseries.py

@@ -952,7 +952,7 @@ def get_top_drawdowns(returns, top=10):
     underwater = df_cum / running_max - 1
 
     drawdowns = []
-    for t in range(top):
+    for _ in range(top):
         peak, valley, recovery = get_max_drawdown_underwater(underwater)
         # Slice out draw-down period
         if not pd.isnull(recovery):

pyfolio/utils.py

@@ -354,7 +354,7 @@ def estimate_intraday(returns, positions, transactions, EOD_hour=23):
     # Calculate exposure, then take peak of exposure every day
     txn_val['exposure'] = txn_val.abs().sum(axis=1)
     condition = (txn_val['exposure'] == txn_val.groupby(
-        pd.TimeGrouper('24H'))['exposure'].transform(max))
+        pd.Grouper(freq='24H'))['exposure'].transform(max))
     txn_val = txn_val[condition].drop('exposure', axis=1)
 
     # Compute cash delta