haojin2 commented on a change in pull request #11587: [MXNET-378] Adding
depth_to_space and space_to_depth operator(Updated)
URL: https://github.com/apache/incubator-mxnet/pull/11587#discussion_r204900419
##########
File path: tests/python/unittest/test_operator.py
##########
@@ -6681,6 +6681,91 @@ def test_diag():
diag_sym = mx.sym.diag(data=data, k=-1)
check_numeric_gradient(diag_sym, [a_np])
+@with_seed()
+def test_depthtospace():
+ def f(x, blocksize):
+ b, c, h, w = x.shape[0], x.shape[1], x.shape[2], x.shape[3]
+ tmp = np.reshape(x, [b, blocksize, blocksize, c // (blocksize**2), h,
w])
+ tmp = np.transpose(tmp, [0, 3, 4, 1, 5, 2])
+ y = np.reshape(tmp, [b, c // (blocksize**2), h * blocksize, w *
blocksize])
+ return y
+
+ shape_inp = (1,4,2,3)
+ block = 2
+ data = rand_ndarray(shape_inp, 'default')
+ data_np = data.asnumpy()
+ expected = f(data_np, block)
+ output = mx.nd.depth_to_space(data, block)
+ assert_almost_equal(output.asnumpy(), expected, atol=1e-3, rtol=1e-3)
+
+ shape_out = (1,1,4,6)
+ data = mx.sym.Variable('data')
+ dts_sym = mx.sym.depth_to_space(data, block)
+ check_numeric_gradient(dts_sym, [np.ones(shape_inp)])
+
+ check_symbolic_forward(dts_sym, [data_np], [expected])
+ check_symbolic_backward(dts_sym, [data_np], [np.ones(shape_out)],
[np.ones(shape_inp)])
+
+ def test_invalid_depth_dim():
+ invalid_shape_inp = (1,3,2,3)
+ block = 2
+ data = rand_ndarray(invalid_shape_inp, 'default')
+ assertRaises(MXNetError, mx.nd.depth_to_space, data, block)
+
+ def test_invalid_space_dim():
+ invalid_shape_inp = (1,4,2,3)
+ block = 2
+ data = rand_ndarray(invalid_shape_inp, 'default')
+ assertRaises(MXNetError, mx.nd.depth_to_space, data, block)
+
+ def test_invalid_block_size():
+ invalid_shape_inp = (1,0,2,3)
+ block = 2
+ data = rand_ndarray(invalid_shape_inp, 'default')
+ assertRaises(MXNetError, mx.nd.depth_to_space, data, block)
+
+@with_seed()
+def test_spacetodepth():
+ def f(x, blocksize):
+ b, c, h, w = x.shape[0], x.shape[1], x.shape[2], x.shape[3]
+ tmp = np.reshape(x, [b, c, h // blocksize, blocksize, w // blocksize,
blocksize])
+ tmp = np.transpose(tmp, [0, 3, 5, 1, 2, 4])
+ y = np.reshape(tmp, [b, c * (blocksize**2), h // blocksize, w //
blocksize])
+ return y
+
+ shape_inp = (1,1,4,6)
+ block = 2
+ data = rand_ndarray(shape_inp, 'default')
+ data_np = data.asnumpy()
+ expected = f(data_np, block)
+ output = mx.nd.space_to_depth(data, block)
+ assert_almost_equal(output.asnumpy(), expected, atol=1e-3, rtol=1e-3)
+
+ shape_out = (1,4,2,3)
Review comment:
Maybe you can randomize the shapes of test inputs to provide more test
coverage. Please do make sure to test for flakiness locally. You can also try
out the latest tool which can help you with that:
https://github.com/apache/incubator-mxnet/blob/master/tools/flakiness_checker.py
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
