qis 3.2.9__tar.gz → 3.2.11__tar.gz

{qis-3.2.9 → qis-3.2.11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: qis
-Version: 3.2.9
+Version: 3.2.11
 Summary: Implementation of visualisation and reporting analytics for Quantitative Investment Strategies
 License: LICENSE.txt
 Keywords: quantitative,investing,portfolio optimization,systematic strategies,volatility

{qis-3.2.9 → qis-3.2.11}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "qis"
-version = "3.2.9"
+version = "3.2.11"
 description = "Implementation of visualisation and reporting analytics for Quantitative Investment Strategies"
 license = "LICENSE.txt"
 authors = ["Artur Sepp <artursepp@gmail.com>"]

{qis-3.2.9 → qis-3.2.11}/qis/examples/factsheets/multi_assets.py

@@ -23,7 +23,7 @@ class UnitTests(Enum):
 
 def run_unit_test(unit_test: UnitTests):
 
-    end_date = '10Jan2025'  # performance repoting
+    end_date = '17Apr2025'  # performance repoting
 
     prices = None  # if Noe, use yahoo finance data
 

{qis-3.2.9 → qis-3.2.11}/qis/plots/lineplot.py

@@ -15,6 +15,9 @@ import qis.plots.utils as put
 
 
 def plot_line(df: Union[pd.Series, pd.DataFrame],
+              x: str = None,
+              y: str = None,
+              hue: str = None,
               linestyles: List[str] = None,
               linestyle: Any = '-',
               linewidth: float = 1.0,
@@ -37,7 +40,13 @@ def plot_line(df: Union[pd.Series, pd.DataFrame],
               ax: plt.Subplot = None,
               **kwargs
               ) -> Optional[plt.Figure]:
-
+    """
+    line plot wrapper for sns, lineplot
+    plot df using index
+    x: str = None: x column
+    y: str = None: y column
+    hue: str = None: hue
+    """
     if isinstance(df, pd.DataFrame):
         pass
     elif isinstance(df, pd.Series):
@@ -45,22 +54,40 @@ def plot_line(df: Union[pd.Series, pd.DataFrame],
     else:
         raise TypeError(f"unsuported data type {type(df)}")
 
+    if x is not None:
+        if y is None:
+            raise ValueError(f"y column must be given with x column")
+        else:
+            if hue is not None:
+                df = df[[x, y, hue]]
+            else:
+                df = df[[x, y]]
+
     if ax is None:
         fig, ax = plt.subplots()
     else:
         fig = None
 
     if colors is None:
-        colors = put.get_n_colors(n=len(df.columns), **kwargs)
+        if hue is None:
+            colors = put.get_n_colors(n=len(df.columns), **kwargs)
+        else:
+            colors = put.get_n_colors(n=len(df[hue].unique()), **kwargs)
 
-    sns.lineplot(data=df, palette=colors, dashes=False, markers=markers, linestyle=linestyle, linewidth=linewidth, ax=ax)
+    sns.lineplot(data=df, x=x, y=y, hue=hue,
+                 palette=colors, dashes=False, markers=markers, linestyle=linestyle, linewidth=linewidth,
+                 style=hue,
+                 ax=ax)
 
     if title is not None:
         put.set_title(ax=ax, title=title, fontsize=fontsize, **kwargs)
 
     if legend_loc is not None:
         if legend_labels is None:
-            legend_labels = put.get_legend_lines(data=df, legend_stats=legend_stats, var_format=yvar_format)
+            if hue is None:
+                legend_labels = put.get_legend_lines(data=df, legend_stats=legend_stats, var_format=yvar_format)
+            else:
+                h, legend_labels = ax.get_legend_handles_labels()
         put.set_legend(ax=ax,
                        labels=legend_labels,
                        colors=colors,

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/multi_portfolio_data.py

@@ -195,7 +195,8 @@ class MultiPortfolioData:
                                                 benchmark_idx: int = 1,
                                                 is_grouped: bool = False,
                                                 group_data: pd.Series = None,
-                                                group_order: List[str] = None
+                                                group_order: List[str] = None,
+                                                total_column: Optional[str] = 'Total'
                                                 ) -> Union[pd.Series, pd.DataFrame]:
         """
         compute Ex ante  tracking error =
@@ -224,7 +225,7 @@ class MultiPortfolioData:
                 group_order = self.portfolio_datas[strategy_idx].group_order
             group_dict = dfg.get_group_dict(group_data=group_data,
                                             group_order=group_order,
-                                            total_column='Total')
+                                            total_column=total_column)
             tracking_error = {key: {} for key in group_dict.keys()}
             for date, pd_covar in self.covar_dict.items():
                 w = weight_diffs.loc[date]
@@ -645,7 +646,7 @@ class MultiPortfolioData:
                              **kwargs)
         if regime_benchmark is not None:
             self.add_regime_shadows(ax=ax, regime_benchmark=regime_benchmark, index=diff.index, regime_params=regime_params)
-    
+
     def get_turnover(self,
                      time_period: TimePeriod = None,
                      turnover_rolling_period: Optional[int] = 12,
@@ -662,7 +663,7 @@ class MultiPortfolioData:
         if time_period is not None:
             turnover = time_period.locate(turnover)
         return turnover
-        
+
     def plot_turnover(self,
                       regime_benchmark: str = None,
                       time_period: TimePeriod = None,
@@ -673,7 +674,7 @@ class MultiPortfolioData:
                       is_unit_based_traded_volume: bool = True,
                       ax: plt.Subplot = None,
                       **kwargs) -> None:
-        
+
         turnover = self.get_turnover(turnover_rolling_period=turnover_rolling_period,
                                      freq_turnover=freq_turnover,
                                      is_unit_based_traded_volume=is_unit_based_traded_volume,
@@ -1039,13 +1040,25 @@ class MultiPortfolioData:
                              regime_params: BenchmarkReturnsQuantileRegimeSpecs = REGIME_PARAMS,
                              time_period: TimePeriod = None,
                              title: Optional[str] = 'Tracking error',
+                             total_column: Optional[str] = 'Total',
                              var_format: str = '{:.2%}',
+                             tre_max_clip: Optional[float] = None,
                              ax: plt.Subplot = None,
                              **kwargs
                              ) -> None:
         tre = self.compute_tracking_error_implied_by_covar(strategy_idx=strategy_idx, benchmark_idx=benchmark_idx,
                                                            is_grouped=is_grouped, group_data=group_data,
-                                                           group_order=group_order)
+                                                           group_order=group_order,
+                                                           total_column=total_column)
+        if tre_max_clip is not None:
+            if isinstance(tre, pd.DataFrame):  # skip total_column from trimming
+                tre_columns = tre.columns.to_list()
+                if total_column in tre:
+                    tre_columns.remove(total_column)
+                tre[tre_columns] = tre[tre_columns].clip(upper=tre_max_clip)
+            else:
+                tre = tre.clip(upper=tre_max_clip)
+
         if time_period is not None:
             tre = time_period.locate(tre)
         pts.plot_time_series(df=tre,

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/portfolio_data.py

@@ -1504,13 +1504,13 @@ class PortfolioData:
         freq = pd.infer_freq(turnover.index)
         turnover_title = title or f"{turnover_rolling_period}-period rolling {freq}-freq Turnover"
         qis.plot_time_series(df=turnover,
-                             var_format='{:,.2%}',
+                             var_format='{:,.1%}',
                              y_limits=(0.0, None),
                              legend_stats=qis.LegendStats.AVG_NONNAN_LAST,
                              title=turnover_title,
                              ax=ax,
                              **kwargs)
-        if regime_benchmark is not None:
+        if regime_benchmark is not None and self.benchmark_prices is not None:
             self.add_regime_shadows(ax=ax, regime_benchmark=regime_benchmark, index=turnover.index,
                                     regime_params=regime_params)
 

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/multi_assets_factsheet.py

@@ -420,7 +420,7 @@ def generate_multi_asset_factsheet(prices: pd.DataFrame,
                                    heatmap_freq: str = 'YE',
                                    time_period: TimePeriod = None,  # time period for reporting
                                    figsize: Tuple[float, float] = (8.3, 11.7),  # A4 for portrait
-                                   fontsize: int = 4,
+                                   fontsize: int = 5,
                                    factsheet_name: str = None,
                                    performance_bars: Tuple[PerfStat, PerfStat] = (PerfStat.SHARPE_RF0, PerfStat.MAX_DD),
                                    drop_1y_ra_perf_table: bool = True,

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/multi_strategy_factseet_pybloqs.py

@@ -232,7 +232,7 @@ def generate_multi_portfolio_factsheet(multi_portfolio_data: MultiPortfolioData,
                                        heatmap_freq: str = 'YE',
                                        figsize: Tuple[float, float] = (8.3, 11.7),  # A4 for portrait
                                        is_grouped: bool = False,
-                                       fontsize: int = 4,
+                                       fontsize: int = 5,
                                        **kwargs
                                        ) -> plt.Figure:
     """

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/multi_strategy_factsheet.py

@@ -27,7 +27,7 @@ def generate_multi_portfolio_factsheet(multi_portfolio_data: MultiPortfolioData,
                                        group_data: pd.Series = None,
                                        add_group_exposures_and_pnl: bool = False,
                                        add_strategy_factsheets: bool = False,
-                                       fontsize: int = 4,
+                                       fontsize: int = 5,
                                        **kwargs
                                        ) -> List[plt.Figure]:
     """

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/strategy_benchmark_factsheet.py

@@ -35,7 +35,7 @@ def generate_strategy_benchmark_factsheet_plt(multi_portfolio_data: MultiPortfol
                                               add_exposures_comp: bool = False,
                                               is_grouped: Optional[bool] = None,
                                               figsize: Tuple[float, float] = (8.3, 11.7),  # A4 for portrait
-                                              fontsize: int = 4,
+                                              fontsize: int = 5,
                                               **kwargs
                                               ) -> List[plt.Figure]:
     """
@@ -513,6 +513,7 @@ def weights_tracking_error_report_by_ac_subac(multi_portfolio_data: MultiPortfol
                                               perf_params: PerfParams = PERF_PARAMS,
                                               regime_params: BenchmarkReturnsQuantileRegimeSpecs = REGIME_PARAMS,
                                               add_benchmarks_to_navs: bool = True,
+                                              tre_max_clip: Optional[float] = None,
                                               figsize: Tuple[float, float] = (11.7, 8.3),
                                               var_format: str = '{:.1%}',
                                               add_titles: bool = True,
@@ -758,6 +759,7 @@ def weights_tracking_error_report_by_ac_subac(multi_portfolio_data: MultiPortfol
                                                   group_order=ac_group_order,
                                                   regime_benchmark=regime_benchmark,
                                                   regime_params=regime_params,
+                                                  tre_max_clip=tre_max_clip,
                                                   title=title,
                                                   ax=ax,
                                                   time_period=time_period,
@@ -785,6 +787,9 @@ def weights_tracking_error_report_by_ac_subac(multi_portfolio_data: MultiPortfol
                                                                          group_order=turnover_order,
                                                                          add_total=False,
                                                                          **kwargs)
+        if regime_benchmark is not None:
+            multi_portfolio_data.add_regime_shadows(ax=ax, regime_benchmark=regime_benchmark,
+                                                    index=grouped_selection_return.index, regime_params=regime_params)
         if not add_titles:
             ax.title.set_visible(False)
     return figs, dfs

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/strategy_benchmark_factsheet_pybloqs.py

@@ -33,7 +33,7 @@ def generate_strategy_benchmark_factsheet_with_pyblogs(multi_portfolio_data: Mul
                                                        heatmap_freq: str = 'YE',
                                                        weight_freq: str = 'ME',
                                                        is_input_weights: bool = True,
-                                                       fontsize: int = 4,
+                                                       fontsize: int = 5,
                                                        **kwargs
                                                        ) -> p.VStack:
     """

{qis-3.2.9 → qis-3.2.11}/qis/portfolio/reports/strategy_factsheet.py

@@ -38,7 +38,7 @@ def generate_strategy_factsheet(portfolio_data: PortfolioData,
                                 sharpe_rolling_window: int = 260,
                                 add_benchmarks_to_navs: bool = False,
                                 figsize: Tuple[float, float] = (8.5, 11.7),  # A4 for portrait
-                                fontsize: int = 4,
+                                fontsize: int = 5,
                                 weight_change_sample_size: int = 20,
                                 add_current_position_var_risk_sheet: bool = False,
                                 add_weights_turnover_sheet: bool = False,

{qis-3.2.9 → qis-3.2.11}/qis/settings.yaml

@@ -18,4 +18,3 @@ OUTPUT_PATH:
 
 AWS_POSTGRES:
   ""
-

{qis-3.2.9 → qis-3.2.11}/qis/utils/dates.py

@@ -28,6 +28,9 @@ CALENDAR_DAYS_IN_MONTH = 30
 CALENDAR_DAYS_PER_YEAR_SHARPE = 365.25  # for total return computations for Sharpe
 
 
+DEFAULT_TRADING_YEAR_DAYS = 252  # How mny trading days we assume per year, see get_period_days()
+
+
 def get_current_time_with_tz(tz: Optional[str] = 'UTC',
                              days_offset: int = None,
                              normalize: bool = True,
@@ -67,7 +70,7 @@ def get_period_days(freq: str = 'B',
     an_f will return the number of period in year
     consistent with using 252 for vol annualization
     """
-    an_days = 365 if is_calendar else 252
+    an_days = 365 if is_calendar else DEFAULT_TRADING_YEAR_DAYS
     if freq in ['1M']:
         days = 1.0 / 24.0 / 60.0
         an_f = an_days * 24.0 * 60.0
@@ -117,7 +120,7 @@ def get_period_days(freq: str = 'B',
         days = an_days
         an_f = 1.0
     else:
-        raise TypeError(f'freq={freq} is not impelemnted')
+        raise TypeError(f'freq={freq} is not implemented')
 
     return days, an_f