trim down tests & add type checking for meshgrid

mindspore/numpy/array_creations.py

@@ -1262,7 +1262,8 @@ def meshgrid(*xi, sparse=False, indexing='xy'):
         along the first dimension for `x1`, the second for `x2` and so on.
 
     Raises:
-        TypeError: if the input is not a tensor.
+        TypeError: if the input is not a tensor, or sparse is not boolean, or
+            indexing is not 'xy' or 'ij'.
 
     Supported Platforms:
         ``Ascend`` ``GPU`` ``CPU``
@@ -1285,6 +1286,10 @@ def meshgrid(*xi, sparse=False, indexing='xy'):
         [1.]
     """
     _check_input_tensor(*xi)
+    if not isinstance(sparse, bool):
+        _raise_type_error('argument sparse should be boolean')
+    if indexing not in ('xy', 'ij'):
+        _raise_type_error("Valid values for `indexing` are 'xy' and 'ij'.")
 
     grids = []
     for x in xi:

tests/st/numpy_native/test_array_creations.py

@@ -744,8 +744,8 @@ def test_diag():
 @pytest.mark.platform_x86_cpu
 @pytest.mark.env_onecard
 def test_diag_indices():
-    mnp_res = mnp.diag_indices(0)
-    onp_res = onp.diag_indices(0)
+    mnp_res = mnp.diag_indices(5, 7)
+    onp_res = onp.diag_indices(5, 7)
     match_all_arrays(mnp_res, onp_res)
 
 

tests/st/numpy_native/test_array_ops.py

@@ -970,7 +970,7 @@ def onp_flip(x):
     return a, b, c, d
 
 
-@pytest.mark.level1
+@pytest.mark.level2
 @pytest.mark.platform_arm_ascend_training
 @pytest.mark.platform_x86_ascend_training
 @pytest.mark.platform_x86_gpu_training
@@ -989,7 +989,7 @@ def onp_flipud(x):
     return  onp.flipud(x)
 
 
-@pytest.mark.level1
+@pytest.mark.level2
 @pytest.mark.platform_arm_ascend_training
 @pytest.mark.platform_x86_ascend_training
 @pytest.mark.platform_x86_gpu_training
@@ -1008,7 +1008,7 @@ def onp_fliplr(x):
     return onp.fliplr(x)
 
 
-@pytest.mark.level1
+@pytest.mark.level2
 @pytest.mark.platform_arm_ascend_training
 @pytest.mark.platform_x86_ascend_training
 @pytest.mark.platform_x86_gpu_training

tests/st/numpy_native/test_math_ops.py

@@ -28,7 +28,6 @@ class Cases():
             rand_int(2),
             rand_int(2, 3),
             rand_int(2, 3, 4),
-            rand_int(2, 3, 4, 5),
         ]
 
         # scalars expanded across the 0th dimension
@@ -36,7 +35,6 @@ class Cases():
             rand_int(),
             rand_int(1),
             rand_int(1, 1),
-            rand_int(1, 1, 1, 1),
         ]
 
         # empty arrays
@@ -44,7 +42,6 @@ class Cases():
             rand_int(0),
             rand_int(4, 0),
             rand_int(2, 0, 2),
-            rand_int(5, 0, 7, 0),
         ]
 
         # arrays of the same size expanded across the 0th dimension
@@ -52,7 +49,6 @@ class Cases():
             rand_int(2, 3),
             rand_int(1, 2, 3),
             rand_int(1, 1, 2, 3),
-            rand_int(1, 1, 1, 2, 3),
         ]
 
         # arrays with last dimension aligned
@@ -68,7 +64,6 @@ class Cases():
             rand_int(5),
             rand_int(6, 1),
             rand_int(7, 1, 5),
-            rand_int(8, 1, 6, 1)
         ]
 
         # boolean arrays which can be broadcast
@@ -972,7 +967,9 @@ def onp_remainder(x, y):
 @pytest.mark.platform_x86_cpu
 @pytest.mark.env_onecard
 def test_remainder():
-    run_binop_test(mnp_remainder, onp_remainder, test_case)
+    x = rand_int(2, 3)
+    y = rand_int(2, 3)
+    match_res(mnp_remainder, onp_remainder, x, y)
 
 
 def mnp_mod(x, y):
@@ -990,7 +987,9 @@ def onp_mod(x, y):
 @pytest.mark.platform_x86_cpu
 @pytest.mark.env_onecard
 def test_mod():
-    run_binop_test(mnp_mod, onp_mod, test_case)
+    x = rand_int(2, 3)
+    y = rand_int(2, 3)
+    match_res(mnp_mod, onp_mod, x, y)
 
 
 def mnp_fmod(x, y):
@@ -1006,7 +1005,9 @@ def onp_fmod(x, y):
 @pytest.mark.platform_x86_cpu
 @pytest.mark.env_onecard
 def test_fmod():
-    run_binop_test(mnp_fmod, onp_fmod, test_case)
+    x = rand_int(2, 3)
+    y = rand_int(2, 3)
+    match_res(mnp_fmod, onp_fmod, x, y)
 
 
 def mnp_fix(x):
@@ -1028,7 +1029,6 @@ def test_fix():
     y = rand_int(2, 3)
     floats = onp.divide(onp.subtract(x, y), y)
     match_res(mnp_fix, onp_fix, floats, error=1e-5)
-    run_binop_test(mnp_fmod, onp_fmod, test_case, error=1e-5)
 
 
 def mnp_trunc(x):