Hidden API for changing solver

docs/EfficientFrontier.rst

@@ -70,8 +70,6 @@ Basic Usage
 .. automodule:: pypfopt.efficient_frontier
 
     .. autoclass:: EfficientFrontier
-        :members:
-        :exclude-members: custom_objective
 
         .. automethod:: __init__
 
@@ -85,6 +83,7 @@ Basic Usage
                 If you want to generate short-only portfolios, there is a quick hack. Multiply
                 your expected returns by -1, then optimise a long-only portfolio.
 
+        .. automethod:: min_volatility
 
         .. automethod:: max_sharpe
 
@@ -110,6 +109,8 @@ Basic Usage
                 :py:meth:`efficient_return`, the optimiser will fail silently and return
                 weird weights. *Caveat emptor* applies!
 
+        .. automethod:: efficient_return
+
         .. automethod:: portfolio_performance
 
             .. tip::
@@ -123,6 +124,12 @@ Basic Usage
                         weights, expected_returns, cov_matrix, verbose=True, risk_free_rate=0.02
                     )
 
+.. note:: 
+
+    PyPortfolioOpt defers to cvxpy's default choice of solver. If you would like to explicitly
+    choose the solver and see verbose output, simply assign ``ef.solver = "ECOS"`` prior to calling
+    the actual optimisation method. You can choose from any of the `supported solvers <https://www.cvxpy.org/tutorial/advanced/index.html#choosing-a-solver>`_.
+
 Adding objectives and constraints
 =================================
 

pypfopt/base_optimizer.py

@@ -113,6 +113,7 @@ class BaseConvexOptimizer(BaseOptimizer):
     - ``n_assets`` - int
     - ``tickers`` - str list
     - ``weights`` - np.ndarray
+    - ``solver`` - str
 
     Public methods:
 
@@ -144,6 +145,8 @@ class BaseConvexOptimizer(BaseOptimizer):
         self._upper_bounds = None
         self._map_bounds_to_constraints(weight_bounds)
 
+        self.solver = None
+
     def _map_bounds_to_constraints(self, test_bounds):
         """
         Process input bounds into a form acceptable by cvxpy and add to the constraints list.
@@ -193,6 +196,10 @@ class BaseConvexOptimizer(BaseOptimizer):
         """
         try:
             opt = cp.Problem(cp.Minimize(self._objective), self._constraints)
+
+            if self.solver is not None:
+                opt.solve(solver=self.solver, verbose=True)
+            else:
                 opt.solve()
         except (TypeError, cp.DCPError):
             raise exceptions.OptimizationError

pypfopt/efficient_frontier.py

@@ -28,7 +28,7 @@ class EfficientFrontier(base_optimizer.BaseConvexOptimizer):
         - ``bounds`` - float tuple OR (float tuple) list
         - ``cov_matrix`` - np.ndarray
         - ``expected_returns`` - np.ndarray
-
+        - ``solver`` - str
 
     - Output: ``weights`` - np.ndarray
 
@@ -265,7 +265,7 @@ class EfficientFrontier(base_optimizer.BaseConvexOptimizer):
         :return: asset weights for the efficient risk portfolio
         :rtype: dict
         """
-        if not isinstance(target_volatility, float) or target_volatility < 0:
+        if not isinstance(target_volatility, (float, int)) or target_volatility < 0:
             raise ValueError("target_volatility should be a positive float")
 
         self._objective = objective_functions.portfolio_return(

tests/test_efficient_frontier.py

@@ -65,6 +65,28 @@ def test_min_volatility():
     )
 
 
+def test_min_volatility_different_solver():
+    ef = setup_efficient_frontier()
+    ef.solver = "ECOS"
+    w = ef.min_volatility()
+    assert isinstance(w, dict)
+    assert set(w.keys()) == set(ef.tickers)
+    np.testing.assert_almost_equal(ef.weights.sum(), 1)
+    assert all([i >= 0 for i in w.values()])
+    test_performance = (0.179312, 0.159151, 1.001015)
+    np.testing.assert_allclose(ef.portfolio_performance(), test_performance, atol=1e-5)
+
+    ef = setup_efficient_frontier()
+    ef.solver = "OSQP"
+    w = ef.min_volatility()
+    np.testing.assert_allclose(ef.portfolio_performance(), test_performance, atol=1e-5)
+
+    ef = setup_efficient_frontier()
+    ef.solver = "SCS"
+    w = ef.min_volatility()
+    np.testing.assert_allclose(ef.portfolio_performance(), test_performance, atol=1e-3)
+
+
 def test_min_volatility_no_rets():
     # Should work with no rets, see issue #82
     df = get_data()