Merge pull request #442 from vikram-narayan/summary_table

Summary table

Author: Vikram Narayan

pyfolio/bayesian.py

@@ -32,7 +32,7 @@
 from empyrical import cum_returns
 
 
-def model_returns_t_alpha_beta(data, bmark, samples=2000):
+def model_returns_t_alpha_beta(data, bmark, samples=2000, progressbar=True):
     """
     Run Bayesian alpha-beta-model with T distributed returns.
 
@@ -86,12 +86,12 @@ def model_returns_t_alpha_beta(data, bmark, samples=2000):
                  mu=mu_reg,
                  sd=sigma,
                  observed=y)
-        trace = pm.sample(samples)
+        trace = pm.sample(samples, progressbar=progressbar)
 
     return model, trace
 
 
-def model_returns_normal(data, samples=500):
+def model_returns_normal(data, samples=500, progressbar=True):
     """
     Run Bayesian model assuming returns are normally distributed.
 
@@ -125,11 +125,11 @@ def model_returns_normal(data, samples=500):
             returns.distribution.variance**.5 *
             np.sqrt(252))
 
-        trace = pm.sample(samples)
+        trace = pm.sample(samples, progressbar=progressbar)
     return model, trace
 
 
-def model_returns_t(data, samples=500):
+def model_returns_t(data, samples=500, progressbar=True):
     """
     Run Bayesian model assuming returns are Student-T distributed.
 
@@ -167,11 +167,11 @@ def model_returns_t(data, samples=500):
                          returns.distribution.variance**.5 *
                          np.sqrt(252))
 
-        trace = pm.sample(samples)
+        trace = pm.sample(samples, progressbar=progressbar)
     return model, trace
 
 
-def model_best(y1, y2, samples=1000):
+def model_best(y1, y2, samples=1000, progressbar=True):
     """
     Bayesian Estimation Supersedes the T-Test
 
@@ -252,7 +252,7 @@ def model_best(y1, y2, samples=1000):
                          returns_group2.distribution.variance**.5 *
                          np.sqrt(252))
 
-        trace = pm.sample(samples)
+        trace = pm.sample(samples, progressbar=progressbar)
     return model, trace
 
 
@@ -347,7 +347,7 @@ def distplot_w_perc(trace, ax):
                ylabel='Belief', yticklabels=[])
 
 
-def model_stoch_vol(data, samples=2000):
+def model_stoch_vol(data, samples=2000, progressbar=True):
     """
     Run stochastic volatility model.
 
@@ -385,7 +385,7 @@ def model_stoch_vol(data, samples=2000):
                                               pm.math.exp(-2 * s))
         StudentT('r', nu, lam=volatility_process, observed=data)
 
-        trace = pm.sample(samples)
+        trace = pm.sample(samples, progressbar=progressbar)
 
     return model, trace
 
@@ -525,7 +525,7 @@ def _plot_bayes_cone(returns_train, returns_test,
 
 
 def run_model(model, returns_train, returns_test=None,
-              bmark=None, samples=500, ppc=False):
+              bmark=None, samples=500, ppc=False, progressbar=True):
     """
     Run one of the Bayesian models.
 
@@ -563,21 +563,27 @@ def run_model(model, returns_train, returns_test=None,
 
     if model == 'alpha_beta':
         model, trace = model_returns_t_alpha_beta(returns_train,
-                                                  bmark, samples)
+                                                  bmark, samples,
+                                                  progressbar=progressbar)
     elif model == 't':
-        model, trace = model_returns_t(returns_train, samples)
+        model, trace = model_returns_t(returns_train, samples,
+                                       progressbar=progressbar)
     elif model == 'normal':
-        model, trace = model_returns_normal(returns_train, samples)
+        model, trace = model_returns_normal(returns_train, samples,
+                                            progressbar=progressbar)
     elif model == 'best':
-        model, trace = model_best(returns_train, returns_test, samples=samples)
+        model, trace = model_best(returns_train, returns_test,
+                                  samples=samples,
+                                  progressbar=progressbar)
     else:
         raise NotImplementedError(
             'Model {} not found.'
             'Use alpha_beta, t, normal, or best.'.format(model))
 
     if ppc:
         ppc_samples = pm.sample_ppc(trace, samples=samples,
-                                    model=model, size=len(returns_test))
+                                    model=model, size=len(returns_test),
+                                    progressbar=progressbar)
         return trace, ppc_samples['returns']
 
     return trace

pyfolio/perf_attrib.py

@@ -14,10 +14,11 @@
 # limitations under the License.
 from __future__ import division
 
+from collections import OrderedDict
 import empyrical as ep
 import pandas as pd
-
 import matplotlib.pyplot as plt
+
 from pyfolio.pos import get_percent_alloc
 from pyfolio.utils import print_table, set_legend_location
 
@@ -124,12 +125,12 @@ def perf_attrib(returns, positions, factor_returns, factor_loadings,
             pd.concat([perf_attrib_by_factor, returns_df], axis='columns'))
 
 
-def create_perf_attrib_stats(perf_attrib):
+def create_perf_attrib_stats(perf_attrib, risk_exposures):
     """
     Takes perf attribution data over a period of time and computes annualized
     multifactor alpha, multifactor sharpe, risk exposures.
     """
-    summary = {}
+    summary = OrderedDict()
     specific_returns = perf_attrib['specific_returns']
     common_returns = perf_attrib['common_returns']
 
@@ -139,6 +140,8 @@ def create_perf_attrib_stats(perf_attrib):
     summary['Multi-factor sharpe'] =\
         ep.sharpe_ratio(specific_returns)
 
+    # empty line between common/specific/total returns
+    summary[' '] = ' '
     summary['Cumulative specific returns'] =\
         ep.cum_returns_final(specific_returns)
     summary['Cumulative common returns'] =\
@@ -147,7 +150,9 @@ def create_perf_attrib_stats(perf_attrib):
         ep.cum_returns_final(perf_attrib['total_returns'])
 
     summary = pd.Series(summary)
-    return summary
+
+    risk_exposure_summary = risk_exposures.sum(axis='rows')
+    return summary, risk_exposure_summary
 
 
 def show_perf_attrib_stats(returns, positions, factor_returns,
@@ -164,46 +169,63 @@ def show_perf_attrib_stats(returns, positions, factor_returns,
         pos_in_dollars=pos_in_dollars,
     )
 
-    perf_attrib_stats = create_perf_attrib_stats(perf_attrib_data)
+    perf_attrib_stats, risk_exposure_stats =\
+        create_perf_attrib_stats(perf_attrib_data, risk_exposures)
+
     print_table(perf_attrib_stats)
-    print_table(risk_exposures)
+    print_table(risk_exposure_stats)
 
 
-def plot_returns(perf_attrib_data, ax=None):
+def plot_returns(perf_attrib_data, cost=None, ax=None):
     """
     Plot total, specific, and common returns.
 
     Parameters
     ----------
     perf_attrib_data : pd.DataFrame
         df with factors, common returns, and specific returns as columns,
-        and datetimes as index
+        and datetimes as index. Assumes the `total_returns` column is NOT
+        cost adjusted.
         - Example:
                         momentum  reversal  common_returns  specific_returns
             dt
             2017-01-01  0.249087  0.935925        1.185012          1.185012
             2017-01-02 -0.003194 -0.400786       -0.403980         -0.403980
 
+    cost : pd.Series, optional
+        if present, gets subtracted from `perf_attrib_data['total_returns']`,
+        and gets plotted separately
+
     ax :  matplotlib.axes.Axes
         axes on which plots are made. if None, current axes will be used
 
     Returns
     -------
     ax :  matplotlib.axes.Axes
     """
+
     if ax is None:
         ax = plt.gca()
 
     returns = perf_attrib_data['total_returns']
+    total_returns_label = 'Total returns'
+
+    if cost is not None:
+        returns = returns - cost
+        total_returns_label += ' (adjusted)'
+
     specific_returns = perf_attrib_data['specific_returns']
     common_returns = perf_attrib_data['common_returns']
 
-    ax.plot(ep.cum_returns(returns), color='g', label='Total returns')
+    ax.plot(ep.cum_returns(returns), color='g', label=total_returns_label)
     ax.plot(ep.cum_returns(specific_returns), color='b',
             label='Cumulative specific returns')
     ax.plot(ep.cum_returns(common_returns), color='r',
             label='Cumulative common returns')
 
+    if cost is not None:
+        ax.plot(cost, color='p', label='Cost')
+
     ax.set_title('Time series of cumulative returns')
     ax.set_ylabel('Returns')
 

pyfolio/tears.py

@@ -1065,7 +1065,8 @@ def create_capacity_tear_sheet(returns, positions, transactions,
 @plotting.customize
 def create_bayesian_tear_sheet(returns, benchmark_rets=None,
                                live_start_date=None, samples=2000,
-                               return_fig=False, stoch_vol=False):
+                               return_fig=False, stoch_vol=False,
+                               progressbar=True):
     """
     Generate a number of Bayesian distributions and a Bayesian
     cone plot of returns.
@@ -1134,14 +1135,16 @@ def create_bayesian_tear_sheet(returns, benchmark_rets=None,
 
     trace_t, ppc_t = bayesian.run_model('t', df_train,
                                         returns_test=df_test,
-                                        samples=samples, ppc=True)
+                                        samples=samples, ppc=True,
+                                        progressbar=progressbar)
     previous_time = timer("T model", previous_time)
 
     # Compute BEST model
     print("\nRunning BEST model")
     trace_best = bayesian.run_model('best', df_train,
                                     returns_test=df_test,
-                                    samples=samples)
+                                    samples=samples,
+                                    progressbar=progressbar)
     previous_time = timer("BEST model", previous_time)
 
     # Plot results
@@ -1213,7 +1216,8 @@ def create_bayesian_tear_sheet(returns, benchmark_rets=None,
     benchmark_rets = benchmark_rets.loc[df_train.index]
     trace_alpha_beta = bayesian.run_model('alpha_beta', df_train,
                                           bmark=benchmark_rets,
-                                          samples=samples)
+                                          samples=samples,
+                                          progressbar=progressbar)
     previous_time = timer("running alpha beta model", previous_time)
 
     # Plot alpha and beta

pyfolio/tests/test_tears.py

@@ -128,4 +128,5 @@ def test_create_bayesian_tear_sheet_breakdown(self, kwargs):
         create_bayesian_tear_sheet(
             self.test_returns,
             live_start_date=self.test_returns.index[-20],
+            progressbar=False,
             **kwargs)