def constant_point_coordinates_test():
"""Test features.ConstantPointCoordinates."""
f = features.ConstantPointCoordinates(strokes=0,
points_per_stroke=2,
pen_down=False)
g = features.ConstantPointCoordinates(strokes=0,
points_per_stroke=200,
pen_down=False)
recording = testhelper.get_symbol_as_handwriting(292934)
f._features_without_strokes(recording)
g._features_without_strokes(recording)
space_evenly = preprocessing.SpaceEvenly()
space_evenly(recording)
f._features_without_strokes(recording)
def test_simple_execution_test():
algorithms = [
preprocessing.RemoveDuplicateTime(),
preprocessing.RemoveDots(),
preprocessing.SpaceEvenly(),
preprocessing.SpaceEvenlyPerStroke(),
preprocessing.DouglasPeucker(),
preprocessing.StrokeConnect(),
preprocessing.DotReduction(),
preprocessing.WildPointFilter(),
preprocessing.WeightedAverageSmoothing(),
]
for algorithm in algorithms:
a = testhelper.get_symbol_as_handwriting(292934)
algorithm(a)
def test_preprocessing_detection_test():
preprocessing_queue = [
{"ScaleAndShift": None},
{"StrokeConnect": None},
{"DouglasPeucker": [{"epsilon": 0.2}]},
{"SpaceEvenly": [{"number": 100}]},
]
correct = [
preprocessing.ScaleAndShift(),
preprocessing.StrokeConnect(),
preprocessing.DouglasPeucker(epsilon=0.2),
preprocessing.SpaceEvenly(number=100),
]
feature_list = preprocessing.get_preprocessing_queue(preprocessing_queue)
# TODO: Not only compare lengths of lists but actual contents.
assert len(feature_list) == len(correct)
def preprocessing_detection_test():
preprocessing_queue = [{
'ScaleAndShift': None
}, {
'StrokeConnect': None
}, {
'DouglasPeucker': [{
'epsilon': 0.2
}]
}, {
'SpaceEvenly': [{
'number': 100
}]
}]
correct = [
preprocessing.ScaleAndShift(),
preprocessing.StrokeConnect(),
preprocessing.DouglasPeucker(epsilon=0.2),
preprocessing.SpaceEvenly(number=100)
]
feature_list = preprocessing.get_preprocessing_queue(preprocessing_queue)
# TODO: Not only compare lengths of lists but actual contents.
nose.tools.assert_equal(len(feature_list), len(correct))