change all (deprecated) np.product occurrences to np.prod

benchmarks/other/sporco/admm/admm.py

@@ -235,7 +235,7 @@ def __init__(self, Nx, yshape, ushape, dtype, opt=None):
         self.Nx = Nx
         # Working variable U has the same dimensionality as constant c
         # in the constraint Ax + By = c
-        self.Nc = np.product(np.array(ushape))
+        self.Nc = np.prod(np.array(ushape))
 
         # DataType option overrides data type inferred from __init__
         # parameters of derived class
@@ -861,7 +861,7 @@ def __init__(self, xshape, dtype, opt=None):
 
         if opt is None:
             opt = ADMMEqual.Options()
-        Nx = np.product(np.array(xshape))
+        Nx = np.prod(np.array(xshape))
         super(ADMMEqual, self).__init__(Nx, xshape, xshape, dtype, opt)
 
 

benchmarks/other/sporco/admm/parcbpdn.py

@@ -103,7 +103,7 @@ def mpraw_as_np(shape, dtype):
       Numpy array
     """
 
-    sz = int(np.product(shape))
+    sz = int(np.prod(shape))
     csz = sz * np.dtype(dtype).itemsize
     raw = mp.RawArray('c', csz)
     return np.frombuffer(raw, dtype=dtype, count=sz).reshape(shape)

benchmarks/other/sporco/dictlrn/prlcnscdl.py

@@ -110,7 +110,7 @@ def mpraw_as_np(shape, dtype):
       Numpy array
     """
 
-    sz = int(np.product(shape))
+    sz = int(np.prod(shape))
     csz = sz * np.dtype(dtype).itemsize
     raw = mp.RawArray('c', csz)
     return np.frombuffer(raw, dtype=dtype, count=sz).reshape(shape)

benchmarks/other/sporco/prox.py

@@ -112,7 +112,7 @@ def ndto2d(x, axis=-1):
     # Reshape into a 2D array with the axes specified by the axis
     # parameter flattened into an index along rows, and the remaining
     # axes flattened into an index aalong the columns
-    xtr = xt.reshape((np.product(xts[0:len(caxis)]), -1))
+    xtr = xt.reshape((np.prod(xts[0:len(caxis)]), -1))
     # Return reshaped array and a tuple containing the information
     # necessary to undo the entire operation
     return xtr, (xts, prm)

dicodile/utils/dictionary.py

@@ -116,7 +116,7 @@ def init_dictionary(X, n_atoms, atom_support, random_state=None):
     X_std = X.std()
     n_channels, *sig_support = X.shape
     valid_support = get_valid_support(sig_support, atom_support)
-    n_patches = np.product(valid_support)
+    n_patches = np.prod(valid_support)
 
     indices = iter(rng.choice(n_patches, size=10 * n_atoms, replace=False))
     D = np.empty(shape=(n_atoms, n_channels, *atom_support))