Fix indicator box

CHANGELOG.md

@@ -1,5 +1,6 @@
 
 * x.x.x
+  - fix bug in IndicatorBox proximal_conjugate
   - allow CCPi Regulariser functions for not CIL object
   - Add norm for CompositionOperator.
   - Refactor SIRT algorithm to make it more computationally and memory efficient

Wrappers/Python/cil/optimisation/functions/IndicatorBox.py

@@ -142,36 +142,6 @@ def __call__(self, x):
             return self.evaluate(x)
         return 0.0
 
-    def proximal_conjugate(self, x, tau, out=None):
-        r'''Proximal operator of the convex conjugate of IndicatorBox at x:
-
-          .. math:: prox_{\tau * f^{*}}(x)
-
-          Parameters
-          ----------
-          x : DataContainer
-              Input to the proximal operator
-          tau : float
-              Step size. Notice it is ignored in IndicatorBox, see ``proximal`` for details
-          out : DataContainer, optional
-              Output of the proximal operator. If not provided, a new DataContainer is created.
-
-        '''
-
-        # x - tau * self.proximal(x/tau, tau)
-        should_return = False
-
-        if out is None:
-            out = self.proximal(x, tau)
-            should_return = True
-        else:
-            self.proximal(x, tau, out=out)
-
-        out.sapyb(-1., x, 1., out=out)
-
-        if should_return:
-            return out
-
     def proximal(self, x, tau, out=None):
         r'''Proximal operator of IndicatorBox at x
 

Wrappers/Python/test/test_functions.py

@@ -848,6 +848,52 @@ def test_Lipschitz4(self):
         f4 = -2 * f2
         assert f4.L == 2 * f2.L
 
+    def test_proximal_conjugate(self):
+        from cil.framework import AcquisitionGeometry, BlockGeometry
+        ag = AcquisitionGeometry.create_Parallel2D()
+        angles = np.linspace(0, 360, 10, dtype=np.float32)
+
+        #default
+        ag.set_angles(angles)
+        ag.set_panel(5)
+
+        ig = ag.get_ImageGeometry()
+        bg = BlockGeometry(ig, ig)
+
+        b = ag.allocate('random', seed=2)
+
+        func_geom_test_list = [
+            (IndicatorBox(), ag),
+            (KullbackLeibler(b=b, backend='numba'), ag),
+            (KullbackLeibler(b=b, backend='numpy'), ag),
+            (L1Norm(), ag),
+            (L2NormSquared(), ag),
+            (MixedL21Norm(), bg),
+            (TotalVariation(backend='c'), ig),
+            (TotalVariation(backend='numpy'), ig),
+        ]
+
+        for func, geom in func_geom_test_list:
+            self.proximal_conjugate_test(func, geom)
+
+    def proximal_conjugate_test(self, function, geom):
+        x = geom.allocate('random', seed=1)
+        tau = 1.0
+        f = Function()
+        f.proximal = function.proximal
+
+        a = function.proximal_conjugate(x, tau)
+        b = f.proximal_conjugate(x, tau)
+
+        # handle the case of MixedL21Norm
+        if isinstance(a, BlockDataContainer):
+            for xa,xb in zip(a,b):
+                np.testing.assert_allclose(xa.as_array(), xb.as_array(),
+                                    rtol=1e-5, atol=1e-5)
+        else:
+            np.testing.assert_allclose(a.as_array(), b.as_array(),
+                                    rtol=1e-5, atol=1e-5)
+
 
 class TestTotalVariation(unittest.TestCase):