Fix cases using arrays instead of scalars

Author: frances-h

copulas/bivariate/base.py

@@ -341,7 +341,7 @@ def partial_derivative_scalar(self, U, V):
         self.check_fit()
 
         X = np.column_stack((U, V))
-        return self.partial_derivative(X)
+        return self.partial_derivative(X).item()
 
     def set_random_state(self, random_state):
         """Set the random state.

copulas/bivariate/frank.py

@@ -166,5 +166,5 @@ def _tau_to_theta(self, alpha):
         def debye(t):
             return t / (np.exp(t) - 1)
 
-        debye_value = integrate.quad(debye, EPSILON, alpha)[0] / alpha
+        debye_value = integrate.quad(debye, EPSILON, alpha.item())[0] / alpha
         return 4 * (debye_value - 1) / alpha + 1 - self.tau

copulas/multivariate/tree.py

@@ -202,8 +202,8 @@ def get_likelihood(self, uni_matrix):
         for i in range(num_edge):
             edge = self.edges[i]
             value, left_u, right_u = edge.get_likelihood(uni_matrix)
-            new_uni_matrix[edge.L, edge.R] = left_u
-            new_uni_matrix[edge.R, edge.L] = right_u
+            new_uni_matrix[edge.L, edge.R] = left_u.item()
+            new_uni_matrix[edge.R, edge.L] = right_u.item()
             values[0, i] = np.log(value)
 
         return np.sum(values), new_uni_matrix

copulas/multivariate/vine.py

@@ -329,7 +329,7 @@ def _sample_row(self):
 
                 new_x = self.ppfs[current](np.array([tmp]))
 
-            sampled[current] = new_x
+            sampled[current] = new_x.item()
 
             for s in neighbors:
                 if s not in visited:

tests/unit/bivariate/test_base.py

@@ -105,12 +105,13 @@ def test_partial_derivative_scalar(self, derivative_mock):
         # Setup
         instance = Bivariate(copula_type=CopulaTypes.CLAYTON)
         instance.fit(self.X)
+        derivative_mock.return_value = np.array([1.0])
 
         # Run
         result = instance.partial_derivative_scalar(0.5, 0.1)
 
         # Check
-        assert result == derivative_mock.return_value
+        assert result == 1.0
 
         expected_args = ((np.array([[0.5, 0.1]]), 0), {})
         assert len(expected_args) == len(derivative_mock.call_args)