Quality of life

In this merge request I propose multiple changes which should improve the usage of the Python API of Aidge:

New Features:

• Add a set_attr method to StaticAttribute objects
• Add a set_attr method to DynamicAttribute objects
• Add magic method __set_attr__ & __get_attr__ to StaticAttribute

Allow the use of

dilation_dims = conv.DilationDims  # __get_attr__ example
conv.DilationDims = [2, 2]         # __set_attr__ example

• Add magic method __set_attr__ & __get_attr__ to DynamicAttribute
• Add **kwargs to GeneriOp ctor to declare more easily attributes:

Usage example

test_op = aidge_core.GenericOperator(
        "Test_generic",
        1,
        0,
        1,
        name="test_op",
        attribute_1=1,
        attribute_2=True,
        attribute_3=0.001
    )

• Allow Numpy->Tensor conversion with no available Backends (move TensorImpl backend_cpu to core)
• Allow to specify a Backend to choose for Numpy->Tensor conversion

aidge_core/unit_tests/test_tensor.py

@@ -10,16 +10,16 @@ SPDX-License-Identifier: EPL-2.0
 
 import unittest
 import aidge_core
-
 from functools import reduce
+
 import numpy as np
 
+
 class test_tensor(unittest.TestCase):
-    """
+    """Test tensor binding
     """
     def setUp(self):
         pass
-
     def tearDown(self):
         pass
 
@@ -35,10 +35,60 @@ class test_tensor(unittest.TestCase):
             idx = t.get_idx(coord)
             self.assertEqual(idx, i)
 
-if __name__ == '__main__':
-    unittest.main()
+    def test_getavailable_backends(self):
+        self.assertTrue("cpu" in aidge_core.Tensor.get_available_backends())
+
+    def test_numpy_int_to_tensor(self):
+        np_array = np.arange(9).reshape(1,1,3,3).astype(np.int32)
+        # Numpy -> Tensor
+        t = aidge_core.Tensor(np_array)
+        self.assertEqual(t.dtype(), aidge_core.DataType.Int32)
+        for i_t, i_n in zip(t, np_array.flatten()):
+            self.assertTrue(i_t == i_n)
+        for i,j in zip(t.dims(), np_array.shape):
+            self.assertEqual(i,j)
+    def test_tensor_int_to_numpy(self):
+        np_array = np.arange(9).reshape(1,1,3,3)
+        # Numpy -> Tensor
+        t = aidge_core.Tensor(np_array)
+        # Tensor -> Numpy
+        nnarray = np.array(t)
+        for i_nn, i_n in zip(nnarray.flatten(), np_array.flatten()):
+            self.assertTrue(i_nn == i_n)
+        for i,j in zip(t.dims(), nnarray.shape):
+            self.assertEqual(i,j)
 
+    def test_numpy_int64_to_tensor(self):
+        np_array = np.arange(9).reshape(1,1,3,3).astype(np.int64)
+        # Numpy -> Tensor
+        t = aidge_core.Tensor(np_array)
+        self.assertEqual(t.dtype(), aidge_core.DataType.Int64)
+        for i_t, i_n in zip(t, np_array.flatten()):
+            self.assertTrue(i_t == i_n)
+        for i,j in zip(t.dims(), np_array.shape):
+            self.assertEqual(i,j)
 
+    def test_numpy_float_to_tensor(self):
+        t = aidge_core.Tensor()
+        np_array = np.random.rand(1, 1, 3, 3).astype(np.float32)
+        # Numpy -> Tensor
+        t = aidge_core.Tensor(np_array)
+        self.assertEqual(t.dtype(), aidge_core.DataType.Float32)
+        for i_t, i_n in zip(t, np_array.flatten()):
+            self.assertTrue(i_t == i_n) # TODO : May need to change this to a difference
+        for i,j in zip(t.dims(), np_array.shape):
+            self.assertEqual(i,j)
 
+    def test_get_set(self):
+        dims = [2,2,2]
 
+        np_array = np.arange(8).reshape(dims).astype(np.int32)
+        # Numpy -> Tensor
+        t = aidge_core.Tensor(np_array)
+        for i in range(8):
+            self.assertEqual(t[i], i)
+            t[i] = 5
+            self.assertEqual(t[i], 5)
 
+if __name__ == '__main__':
+    unittest.main()