def upsample2_test(self):
"""
Here testing upsampling with the following configurations (+ testing stereo):
IN: | | | |
OUT: | | | | |
"""
def gen():
"""
[[0, 0], [-2, 2], [-4, 4]] [[-6, 6], [-8, 8], [-10, 10]] ...
"""
for i in range(0, 2):
block_in = numpy.vstack([
numpy.arange(-i * 2 * 3, -(i + 1) * 2 * 3, -2),
numpy.arange(i * 2 * 3, (i + 1) * 2 * 3, 2)
]).transpose()
yield block_in
resampler = stream_functions.resample(gen())
ratio = 3.0 / 2
resampler.set_ratio(ratio)
# IN: 0 1 2 3 4 5
# OUT: 0 1 2 3 4 5 6 7
numpy.testing.assert_array_equal(
stream_functions.concatenate(resampler).round(8),
(numpy.array([
[0, 0], [-1, 1], [-2, 2], [-3, 3],
[-4, 4], [-5, 5], [-6, 6], [-7, 7]
]) * 2 / ratio).round(8)
)
def upsample1_test(self):
"""
Testing upsampling with the following configurations :
IN: | | |
OUT: | | | | | | |
"""
def gen():
"""
[[0], [2], [4]] [[6], [8], [10]] ...
"""
for i in range(0, 2):
yield numpy.arange(i * 2 * 3, (i + 1) * 2 * 3, 2).reshape(3, 1)
resampler = stream_functions.resample(gen())
resampler.set_ratio(3.0)
# IN: 0 1 2 3 4 5
# OUT: 0 1 2 3 4 5 6 7 8 9 a b c d e f
numpy.testing.assert_array_equal(
stream_functions.concatenate(resampler).round(8),
(numpy.array([
[0], [1], [2], [3], [4], [5], [6], [7], [8],
[9], [10], [11], [12], [13], [14], [15]
]) * 2.0 / 3).round(8)
)
def simple_test(self):
def source():
for i in range(0, 3):
yield numpy.ones([3, 1]) * i
block = stream_functions.concatenate(source())
numpy.testing.assert_array_equal(block, numpy.array([
[0], [0], [0], [1], [1], [1], [2], [2], [2]
]))
def hop_size_bigger_than_win_size_test(self):
def gen():
for i in range(6):
yield numpy.array([[i * 11, i * 11]])
window = stream_functions.window(gen(), 2, 3, pad=True)
numpy.testing.assert_array_equal(
stream_functions.concatenate(window),
[[0, 0], [11, 11], [33, 33], [44, 44]]
)
def win_size_exact_and_pad_test(self):
"""
Test when padding is True, and last window falls exactly, without actual need for padding.
"""
def gen():
for i in range(2):
yield numpy.array([[i * 11, i * 11]])
window = stream_functions.window(gen(), 1, 1, pad=True)
numpy.testing.assert_array_equal(
stream_functions.concatenate(window),
[[0, 0], [11, 11]]
)
def ratio1_test(self):
"""
Ratio 1 test.
"""
def gen():
"""
[[0], [0.5], [1]] [[1.5], [2], [2.5]] ...
"""
for i in range(0, 2):
yield numpy.arange(i * 3 * 0.5, (i + 1) * 3 * 0.5, 0.5).reshape(3, 1)
resampler = stream_functions.resample(gen())
numpy.testing.assert_array_equal(
stream_functions.concatenate(resampler).round(8),
(numpy.array([[0], [1], [2], [3], [4], [5]]) * 0.5).round(8)
)
def downsample3_test(self):
"""
Testing high downsampling, several blocks of incoming data fetched for one frame out.
"""
def gen():
"""
[[0], [0.5], [1]] [[1.5], [2], [2.5]] ...
"""
for i in range(0, 6):
yield numpy.arange(i * 3 * 0.5, (i + 1) * 3 * 0.5, 0.5).reshape(3, 1)
resampler = stream_functions.resample(gen())
ratio = 1.0 / 8
resampler.set_ratio(ratio)
# IN: 0 1 2 3 4 5 6 7 8 9 a b c d e f g h
# OUT: 0 1 2
numpy.testing.assert_array_equal(
stream_functions.concatenate(resampler).round(8),
(numpy.array([[0], [1], [2]]) * (0.5 / ratio)).round(8)
)
def downsample2_test(self):
"""
Testing downsampling with the following configurations:
# IN: | | | | | | | | |
# OUT: | | |
"""
def gen():
"""
[[0], [0.5], [1]] [[1.5], [2], [2.5]] ...
"""
for i in range(0, 3):
yield numpy.arange(i * 3 * 0.5, (i + 1) * 3 * 0.5, 0.5).reshape(3, 1)
resampler = stream_functions.resample(gen())
ratio = 1.0 / 4
resampler.set_ratio(ratio)
# IN: 0 1 2 3 4 5 6 7 8
# OUT: 0 1 2
numpy.testing.assert_array_equal(
stream_functions.concatenate(resampler).round(8),
(numpy.array([[0], [1], [2]]) * 0.5 / ratio).round(8)
)
