revert the wrong refacor of iteritems of series

add cap_buckets & sector_dict into tear function

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].items():
+        for stat, value in perf_stats[column].iteritems():
             if stat in STAT_FUNCS_PCT:
                 perf_stats.loc[stat, column] = str(np.round(value * 100,
                                                             1)) + '%'

pyfolio/risk.py

@@ -133,9 +133,7 @@ def compute_sector_exposures(positions, sectors, sector_dict=None):
 
     sector_dict : dict or OrderedDict
         Dictionary of all sectors
-        - Keys are sector codes (e.g. ints or strings) and values are sector
-          names (which must be strings)
-        - Defaults to Morningstar sectors
+        - See full explanation in create_risk_tear_sheet
     """
 
     if sector_dict is None:
@@ -282,7 +280,7 @@ def plot_sector_exposures_net(net_exposures, sector_dict=None, ax=None):
     return ax
 
 
-def compute_cap_exposures(positions, caps):
+def compute_cap_exposures(positions, caps, cap_buckets=None):
     """
     Returns arrays of long, short and gross market cap exposures of an
     algorithm's positions
@@ -296,8 +294,16 @@ def compute_cap_exposures(positions, caps):
     caps : pd.DataFrame
         Daily Morningstar sector code per asset
         - See full explanation in create_risk_tear_sheet
+
+    cap_buckets : dict or OrderedDict
+        Dictionary of all cap buckets
+        - See full explanation in create_risk_tear_sheet
+
     """
 
+    if cap_buckets is None:
+        cap_buckets = CAP_BUCKETS
+
     long_exposures = []
     short_exposures = []
     gross_exposures = []
@@ -310,7 +316,7 @@ def compute_cap_exposures(positions, caps):
     tot_short_exposure = positions_wo_cash[positions_wo_cash < 0] \
         .abs().sum(axis='columns')
 
-    for bucket_name, boundaries in CAP_BUCKETS.items():
+    for bucket_name, boundaries in cap_buckets.items():
         in_bucket = positions_wo_cash[(caps >= boundaries[0]) &
                                       (caps <= boundaries[1])]
 

pyfolio/tears.py

@@ -1062,7 +1062,9 @@ def create_capacity_tear_sheet(returns, positions, transactions,
 def create_risk_tear_sheet(positions,
                            style_factors,
                            sectors=None,
+                           sector_dict=None,
                            caps=None,
+                           cap_buckets=None,
                            shares_held=None,
                            volumes=None,
                            percentile=None,
@@ -1110,6 +1112,12 @@ def create_risk_tear_sheet(positions,
         2017-04-04	     311.0       206.0
         2017-04-05	     311.0	     206.0
 
+    sector_dict : dict or OrderedDict
+        Dictionary of all sectors
+        - Keys are sector codes (e.g. ints or strings) and values are sector
+          names (which must be strings)
+        - Defaults to Morningstar sectors, see SECTORS in risk.py
+
     caps : pd.DataFrame
         Daily market cap per asset
         - DataFrame with dates as index and equities as columns
@@ -1120,6 +1128,12 @@ def create_risk_tear_sheet(positions,
         2017-04-04	   1.329620e+10     6.403694e+10
         2017-04-05	   1.297464e+10	    6.397187e+10
 
+    cap_buckets : dict or OrderedDict
+        Dictionary of all cap buckets
+        - Keys are Micro/Small/Mid/Large/Mega
+        - Values are capital value range tuple
+        - Defaults to CAP_BUCKETS in risk.py
+
     shares_held : pd.DataFrame
         Daily number of shares held by an algorithm.
         - Example:
@@ -1199,7 +1213,7 @@ def create_risk_tear_sheet(positions,
         i += 1
         ax_sector_net = plt.subplot(gs[i, :], sharex=style_axes[0])
         long_exposures, short_exposures, gross_exposures, net_exposures \
-            = risk.compute_sector_exposures(positions, sectors)
+            = risk.compute_sector_exposures(positions, sectors, sector_dict)
         risk.plot_sector_exposures_longshort(long_exposures, short_exposures,
                                              ax=ax_sector_longshort)
         risk.plot_sector_exposures_gross(gross_exposures, ax=ax_sector_gross)
@@ -1213,7 +1227,7 @@ def create_risk_tear_sheet(positions,
         i += 1
         ax_cap_net = plt.subplot(gs[i, :], sharex=style_axes[0])
         long_exposures, short_exposures, gross_exposures, net_exposures \
-            = risk.compute_cap_exposures(positions, caps)
+            = risk.compute_cap_exposures(positions, caps, cap_buckets)
         risk.plot_cap_exposures_longshort(long_exposures, short_exposures,
                                           ax_cap_longshort)
         risk.plot_cap_exposures_gross(gross_exposures, ax_cap_gross)