Merge pull request #21 from skfolio/fix/issue-19/number-of-assets

[BUG] Number of Assets and Portfolio len

src/skfolio/measures/_measures.py

@@ -614,7 +614,7 @@ def effective_number_assets(weights: np.ndarray) -> float:
     r"""
     Computes the effective number of assets, defined as the inverse of the Herfindahl index [1]_:
     .. math:: N_{eff} = \frac{1}{\Vert w \Vert_{2}^{2}}
-    
+
     It quantifies portfolio concentration, with a higher value indicating a more diversified portfolio.
 
     Parameters
@@ -632,4 +632,4 @@ def effective_number_assets(weights: np.ndarray) -> float:
     .. [1] "Banking and Financial Institutions Law in a Nutshell".
         Lovett, William Anthony (1988)
     """
-    return 1.0/(np.power(weights, 2).sum())
+    return 1.0 / (np.power(weights, 2).sum())

src/skfolio/portfolio/_base.py

@@ -369,7 +369,6 @@ class BasePortfolio:
     _read_only_attrs: ClassVar[set] = {
         "returns",
         "observations",
-        "n_observations",
     }
 
     # Arguments globally used in measures computation
@@ -401,7 +400,6 @@ class BasePortfolio:
         # public read-only
         "returns",
         "observations",
-        "n_observations",
         # private
         "_loaded",
         # custom getter and setter
@@ -521,7 +519,6 @@ def __init__(
             self._fitness_measures = [PerfMeasure.MEAN, RiskMeasure.VARIANCE]
         else:
             self._fitness_measures = fitness_measures
-        self.n_observations = len(observations)
         self._loaded = True
 
     def __reduce__(self):
@@ -531,9 +528,6 @@ def __reduce__(self):
             [getattr(self, arg) for arg in args_names(self.__init__)]
         )
 
-    def __len__(self) -> int:
-        return len(self.observations)
-
     def __repr__(self) -> str:
         return f"<{type(self).__name__} {self.name}>"
 
@@ -679,6 +673,11 @@ def drawdowns(self) -> np.ndarray:
         return mt.get_drawdowns(returns=self.returns, compounded=self.compounded)
 
     # Classic property
+    @property
+    def n_observations(self) -> int:
+        """Number of observations"""
+        return len(self.observations)
+
     @property
     def returns_df(self) -> pd.Series:
         """Portfolio returns DataFrame."""

src/skfolio/portfolio/_multi_period_portfolio.py

@@ -369,6 +369,9 @@ def __init__(
         self.check_observations_order = check_observations_order
         self._set_portfolios(portfolios=portfolios)
 
+    def __len__(self) -> int:
+        return len(self.portfolios)
+
     def __getitem__(self, key: int | slice) -> Portfolio | list[Portfolio]:
         return self._portfolios[key]
 
@@ -571,12 +574,12 @@ def summary(self, formatted: bool = True) -> pd.Series:
         """
         df = super().summary(formatted=formatted)
         portfolios_number = len(self)
-        avg_assets_per_portfolio = np.mean([len(p) for p in self])
+        avg_assets_per_portfolio = np.mean([p.n_assets for p in self])
         if formatted:
             portfolios_number = str(int(portfolios_number))
             avg_assets_per_portfolio = f"{avg_assets_per_portfolio:0.1f}"
-        df["Portfolios number"] = portfolios_number
-        df["Avg nb of assets per portfolio"] = avg_assets_per_portfolio
+        df["Portfolios Number"] = portfolios_number
+        df["Avg nb of Assets per Portfolio"] = avg_assets_per_portfolio
         return df
 
     # Public methods

src/skfolio/utils/fixes/_dendrogram.py

@@ -290,15 +290,13 @@ def set_figure_layout(self, width, height):
         Sets and returns default layout object for dendrogram figure.
 
         """
-        self.layout.update(
-            {
-                "showlegend": False,
-                "autosize": False,
-                "hovermode": "closest",
-                "width": width,
-                "height": height,
-            }
-        )
+        self.layout.update({
+            "showlegend": False,
+            "autosize": False,
+            "hovermode": "closest",
+            "width": width,
+            "height": height,
+        })
 
         self.set_axis_layout(self.xaxis)
         self.set_axis_layout(self.yaxis)

tests/test_model_selection/test_validation.py

@@ -62,4 +62,5 @@ def test_validation_combinatorial(X):
     for p in pred:
         assert isinstance(p, MultiPeriodPortfolio)
         assert len(p.portfolios) == cv.n_folds
-        assert len(p) == n_observations
+        assert len(p) == cv.n_folds
+        assert p.n_observations == n_observations

tests/test_portfolio/test_multi_period_portfolio.py

@@ -203,7 +203,8 @@ def test_portfolio_annualized(portfolio_and_returns, annualized_factor):
 
 def test_portfolio_methods(portfolio_and_returns, periods):
     portfolio, returns = portfolio_and_returns
-    assert len(portfolio) == returns.shape[0]
+    assert portfolio.n_observations == returns.shape[0]
+    assert len(portfolio) == len(periods)
     np.testing.assert_almost_equal(returns, portfolio.returns)
     np.testing.assert_almost_equal(returns.mean(), portfolio.mean)
     np.testing.assert_almost_equal(returns.std(ddof=1), portfolio.standard_deviation)
@@ -386,3 +387,10 @@ def test_portfolio_delete_attr(portfolio_and_returns, periods):
         raise
     except AttributeError as e:
         assert str(e) == "`MultiPeriodPortfolio` object has no attribute 'dummy'"
+
+
+def test_portfolio_summary(portfolio_and_returns, periods):
+    portfolio, returns = portfolio_and_returns
+    df = portfolio.summary(formatted=False)
+    assert df.loc["Portfolios Number"] == 3.0
+    assert df.loc["Avg nb of Assets per Portfolio"] == 20.0

tests/test_portfolio/test_portfolio.py

@@ -174,7 +174,8 @@ def test_portfolio_annualized(X, weights, annualized_factor):
 def test_portfolio_methods(X, weights):
     portfolio = Portfolio(X=X, weights=weights)
     returns = _portfolio_returns(asset_returns=X.to_numpy(), weights=weights)
-    assert len(portfolio) == X.shape[0]
+    assert portfolio.n_observations == X.shape[0]
+    assert portfolio.n_assets == X.shape[1]
     np.testing.assert_almost_equal(returns, portfolio.returns)
     np.testing.assert_almost_equal(returns.mean(), portfolio.mean)
     np.testing.assert_almost_equal(returns.std(ddof=1), portfolio.standard_deviation)
@@ -235,7 +236,8 @@ def test_portfolio_magic_methods(X, weights):
     n_assets = X.shape[1]
     ptf_1 = Portfolio(X=X, weights=rand_weights(n=n_assets))
     ptf_2 = Portfolio(X=X, weights=rand_weights(n=n_assets))
-    assert len(ptf_1) == X.shape[0]
+    assert ptf_1.n_observations == X.shape[0]
+    assert ptf_1.n_assets == X.shape[1]
     ptf = ptf_1 + ptf_2
     assert np.array_equal(ptf.weights, ptf_1.weights + ptf_2.weights)
     ptf = ptf_1 - ptf_2