def test_normal_plate():
# 384
plate_384 = plate_generator.normal_plate(plate=384)
assert plate_384.shape == SHAPE_384
assert isinstance(plate_384.data, np.ndarray)
# 1536
plate_1536 = plate_generator.normal_plate(plate=1536)
assert plate_1536.shape == SHAPE_1536
assert isinstance(plate_1536.data, np.ndarray)
def test_normalise_inplace():
plate = plate_generator.normal_plate()
plate = plate + 10
# check the inital plate isn't normalised by accident
assert abs(plate.data.mean()) > EPS
assert abs(plate.data.std() - 1) > EPS
plate._normalise()
# check the normalised plate has a mean of 0 and std of 1
assert abs(plate.data.mean()) < EPS
assert abs(plate.data.std() - 1) < EPS
def test_divide_plates():
plate_a = plate_generator.normal_plate()
plate_b = plate_generator.column_plate()
div_plate = plate_a / plate_b
assert isinstance(div_plate, plate_generator.Plate)
# do the same operation in numpy
div_array = plate_a.data / plate_b.data
assert (div_array == div_plate.data).all()
# try adding a scalar value
div_two = plate_a / 2
assert isinstance(div_two, plate_generator.Plate)
assert isinstance(div_two.data, np.ndarray)
# to the same operation in numpy
div_two_array = plate_a.data / 2
assert (div_two.data == div_two_array).all()
def test_subtract_plates():
plate_a = plate_generator.normal_plate()
plate_b = plate_generator.column_plate()
subtracted_plate = plate_a - plate_b
assert isinstance(subtracted_plate, plate_generator.Plate)
# do the same operation in numpy
subtracted_array = plate_a.data - plate_b.data
assert (subtracted_array == subtracted_plate.data).all()
# try adding a scalar value
subtract_one = plate_a - 1
assert isinstance(subtract_one, plate_generator.Plate)
assert isinstance(subtract_one.data, np.ndarray)
# to the same operation in numpy
subtract_one_array = plate_a.data - 1
assert (subtract_one.data == subtract_one_array).all()
def test_multiply_plates():
plate_a = plate_generator.normal_plate()
plate_b = plate_generator.column_plate()
combined_plate = plate_a * plate_b
assert isinstance(combined_plate, plate_generator.Plate)
# do the same operation in numpy
combined_array = plate_a.data * plate_b.data
assert (combined_array == combined_plate.data).all()
# try adding a scalar value
mult_two = plate_a * 2
assert isinstance(mult_two, plate_generator.Plate)
assert isinstance(mult_two.data, np.ndarray)
# to the same operation in numpy
mult_two_array = plate_a.data * 2
assert (mult_two.data == mult_two_array).all()
def test_add_plates():
plate_a = plate_generator.normal_plate()
plate_b = plate_generator.column_plate()
combined_plate = plate_a + plate_b
assert isinstance(combined_plate, plate_generator.Plate)
# do the same operation in numpy
combined_array = plate_a.data + plate_b.data
assert (combined_array == combined_plate.data).all()
# try adding a scalar value
add_one = plate_a + 1
assert isinstance(add_one, plate_generator.Plate)
assert isinstance(add_one.data, np.ndarray)
# to the same operation in numpy
add_one_array = plate_a.data + 1
assert (add_one.data == add_one_array).all()
# add multiple
plate_c = plate_generator.column_plate()
combined_three = plate_a + plate_b + plate_c
assert isinstance(combined_three.data, np.ndarray)
assert isinstance(combined_three, plate_generator.Plate)
def test_operations_are_commutative():
"""e.g plate + 1 == 1 + plate"""
eps = 1e-6
my_plate = plate_generator.normal_plate()
# should run without error
add_plate_0 = my_plate + 1
add_plate_1 = 1 + my_plate
assert (abs(add_plate_0.data - add_plate_1.data) < eps).all()
array = np.random.normal(size=my_plate.shape)
add_plate_2 = my_plate + array
add_plate_3 = array + my_plate
assert (abs(add_plate_2.data - add_plate_3.data) < eps).all()
sub_plate_0 = my_plate - 1
sub_plate_1 = 1 - my_plate
assert (abs(sub_plate_0.data - sub_plate_1.data) < eps).all()
array_ones = np.ones(my_plate.shape)
sub_plate_2 = my_plate - array_ones
sub_plate_3 = array_ones - my_plate
assert (abs(sub_plate_2.data - sub_plate_3.data) < eps).all()
mul_plate_0 = my_plate * 3
mul_plate_1 = 3 * my_plate
assert (abs(mul_plate_0.data - mul_plate_1.data) < eps).all()
mul_array = np.full(my_plate.shape, 3)
mul_plate_2 = my_plate * mul_array
mul_plate_3 = mul_array * my_plate
assert (abs(mul_plate_2.data - mul_plate_3.data) < eps).all()
div_plate_0 = my_plate / 10.0
div_plate_1 = 10.0 / my_plate
assert (abs(div_plate_0.data - div_plate_1.data) < eps).all()
div_array = np.full_like(my_plate, 10.0)
div_plate_2 = my_plate / div_array
div_plate_3 = div_array / my_plate
assert (abs(div_plate_2.data - div_plate_3.data) < eps).all()
floor_div_plate_0 = my_plate // 2
floor_div_plate_1 = 2 // my_plate
assert (abs(floor_div_plate_0.data - floor_div_plate_1.data) < eps).all()
def test_raises_TypeError():
plate = plate_generator.normal_plate()
with pytest.raises(TypeError):
tmp = plate + "string"
tmp = plate + [1, 2, 3, 4]
tmp = plate + (1, 2, 3, 4)
def test_std():
plate = plate_generator.normal_plate()
output = plate.std()
np_output = plate.data.std()
assert isinstance(output, float)
assert output == np_output