Features/240 flipud

CHANGELOG.md

@@ -5,6 +5,7 @@
 - [#483](https://github.com/helmholtz-analytics/heat/pull/483) Bugfix: Underlying torch tensor moves to the right device on array initialisation
 - [#483](https://github.com/helmholtz-analytics/heat/pull/483) Bugfix: DNDarray.cpu() changes heat device to cpu
 - Update documentation theme to "Read the Docs"
+- [#496](https://github.com/helmholtz-analytics/heat/pull/496) New feature: flipud()
 - [#498](https://github.com/helmholtz-analytics/heat/pull/498) Feature: flip()
 - [#499](https://github.com/helmholtz-analytics/heat/pull/499) Bugfix: MPI datatype mapping: `torch.int16` now maps to `MPI.SHORT` instead of `MPI.SHORT_INT`
 - [#515] (https://github.com/helmholtz-analytics/heat/pull/515) ht.var() now returns the unadjusted sample variance by default, Bessel's correction can be applied by setting ddof=1.

heat/core/manipulations.py

@@ -15,6 +15,7 @@
     "diagonal",
     "expand_dims",
     "flip",
+    "flipud",
     "hstack",
     "resplit",
     "sort",
@@ -619,6 +620,35 @@ def flip(a, axis=None):
     return res
 
 
+def flipud(a):
+    """
+        Flip array in the up/down direction.
+
+        Parameters
+        ----------
+        a: ht.DNDarray
+            Input array to be flipped
+
+        Returns
+        -------
+        res: ht.DNDarray
+            The flipped array.
+
+        Examples
+        --------
+        >>> a = ht.array([[0,1],[2,3]])
+        >>> ht.flipud(a)
+        tensor([[2, 3],
+            [0, 1]])
+
+        >>> b = ht.array([[0,1,2],[3,4,5]], split=0)
+        >>> ht.flipud(b)
+        (1/2) tensor([3,4,5])
+        (2/2) tensor([0,1,2])
+    """
+    return flip(a, 0)
+
+
 def hstack(tup):
     """
     Stack arrays in sequence horizontally (column wise).

heat/core/tests/test_manipulations.py

@@ -786,6 +786,44 @@ def test_flip(self):
         )
         self.assertTrue(ht.equal(ht.flip(a, [1, 2]), r_a))
 
+    def test_flipud(self):
+        a = ht.array([1, 2], device=ht_device)
+        r_a = ht.array([2, 1], device=ht_device)
+        self.assertTrue(ht.equal(ht.flipud(a), r_a))
+
+        b = ht.array([[1, 2], [3, 4]], device=ht_device)
+        r_b = ht.array([[3, 4], [1, 2]], device=ht_device)
+        self.assertTrue(ht.equal(ht.flipud(b), r_b))
+
+        # splitted
+        c = ht.array(
+            [[[0, 1], [2, 3]], [[4, 5], [6, 7]], [[8, 9], [10, 11]], [[12, 13], [14, 15]]],
+            split=0,
+            device=ht_device,
+        )
+        r_c = ht.array(
+            [[[12, 13], [14, 15]], [[8, 9], [10, 11]], [[4, 5], [6, 7]], [[0, 1], [2, 3]]],
+            split=0,
+            device=ht_device,
+        )
+        self.assertTrue(ht.equal(ht.flipud(c), r_c))
+
+        c = ht.array(
+            [[[0, 1], [2, 3]], [[4, 5], [6, 7]], [[8, 9], [10, 11]], [[12, 13], [14, 15]]],
+            split=1,
+            device=ht_device,
+            dtype=ht.float32,
+        )
+        self.assertTrue(ht.equal(ht.resplit(ht.flipud(c), 0), r_c))
+
+        c = ht.array(
+            [[[0, 1], [2, 3]], [[4, 5], [6, 7]], [[8, 9], [10, 11]], [[12, 13], [14, 15]]],
+            split=2,
+            device=ht_device,
+            dtype=ht.int8,
+        )
+        self.assertTrue(ht.equal(ht.resplit(ht.flipud(c), 0), r_c))
+
     def test_hstack(self):
         # cases to test:
         # MM===================================