def func(dtype):
d1 = distance(shape, center=center, scale=scale, dtype=dtype)
d2 = _distance_slow(shape, center, scale, dtype)
assert_same(d1, ref)
assert_same(d2, ref)
distance(shape, center=center, scale=scale, dtype=dtype, out=d1)
_distance_slow(shape, center, scale, dtype, d2)
assert_same(d1, ref)
assert_same(d2, ref)
def func(dtype):
d1 = distance(shape, center=center, scale=scale, dtype=dtype)
d2 = _distance_slow(shape, center, scale, dtype)
assert_same(d1, ref)
assert_same(d2, ref)
distance(shape, center=center, scale=scale, dtype=dtype, out=d1)
_distance_slow(shape, center, scale, dtype, d2)
assert_same(d1, ref)
assert_same(d2, ref)
def integrated_profile_slow(input, center=None, scale=1, bin=1):
d = distance(input.shape, center=center, scale=scale)
m = int(np.max(d))
x = np.ndarray(m + 1)
y = np.ndarray(m + 1)
for i in range(m + 1):
x[i] = (i + 1) * bin
y[i] = np.sum(input[d < x[i]])
return x * scale, y * scale**2
def integrated_profile_slow(input, center=None, scale=1, bin=1):
d = distance(input.shape, center=center, scale=scale)
m = int(np.max(d))
x = np.ndarray(m + 1)
y = np.ndarray(m + 1)
for i in range(m + 1):
x[i] = (i + 1) * bin
y[i] = np.sum(input[d < x[i]])
return x * scale, y * scale**2
def profile_slow(input, center=None, scale=1., bin=1.):
d = distance(input.shape, center=center, scale=scale)
d /= bin
d = d.astype(int)
m = int(np.max(d))
p = np.ndarray(m + 1)
n = np.zeros(m + 1, int)
for i in range(m + 1):
mask = d == i
p[i] = np.mean(input[mask])
n[i] = np.sum(mask)
return p, n
def profile_slow(input, center=None, scale=1., bin=1.):
d = distance(input.shape, center=center, scale=scale)
d /= bin
d = d.astype(int)
m = int(np.max(d))
p = np.ndarray(m + 1)
n = np.zeros(m + 1, int)
for i in range(m + 1):
mask = d == i
p[i] = np.mean(input[mask])
n[i] = np.sum(mask)
return p, n
def test_integrated_profile():
def integrated_profile_slow(input, center=None, scale=1, bin=1):
d = distance(input.shape, center=center, scale=scale)
m = int(np.max(d))
x = np.ndarray(m + 1)
y = np.ndarray(m + 1)
for i in range(m + 1):
x[i] = (i + 1) * bin
y[i] = np.sum(input[d < x[i]])
return x * scale, y * scale**2
d = distance((10, 20))
bin = 2
center = (4, 5)
scale = 0.4
x, y = integrated_profile(d, bin=bin, center=center, scale=scale)
x_ref, y_ref = integrated_profile_slow(d, bin=bin, center=center,
scale=scale)
assert_same(x, x_ref[0:y.size])
assert_same(y, y_ref[0:y.size])
def test_profile():
def profile_slow(input, center=None, scale=1., bin=1.):
d = distance(input.shape, center=center, scale=scale)
d /= bin
d = d.astype(int)
m = int(np.max(d))
p = np.ndarray(m + 1)
n = np.zeros(m + 1, int)
for i in range(m + 1):
mask = d == i
p[i] = np.mean(input[mask])
n[i] = np.sum(mask)
return p, n
d = distance((10, 20))
center = (4, 5)
scale = 0.4
bin = 2
x, y, n = profile(d, center=center, scale=scale, bin=bin, histogram=True)
y_ref, n_ref = profile_slow(d, center=center, scale=scale, bin=bin)
assert_same(y, y_ref[0:y.size])
assert_same(n, n_ref[0:n.size])
def test_integrated_profile():
def integrated_profile_slow(input, center=None, scale=1, bin=1):
d = distance(input.shape, center=center, scale=scale)
m = int(np.max(d))
x = np.ndarray(m + 1)
y = np.ndarray(m + 1)
for i in range(m + 1):
x[i] = (i + 1) * bin
y[i] = np.sum(input[d < x[i]])
return x * scale, y * scale**2
d = distance((10, 20))
bin = 2
center = (4, 5)
scale = 0.4
x, y = integrated_profile(d, bin=bin, center=center, scale=scale)
x_ref, y_ref = integrated_profile_slow(d,
bin=bin,
center=center,
scale=scale)
assert_same(x, x_ref[0:y.size])
assert_same(y, y_ref[0:y.size])
def test_profile():
def profile_slow(input, center=None, scale=1., bin=1.):
d = distance(input.shape, center=center, scale=scale)
d /= bin
d = d.astype(int)
m = int(np.max(d))
p = np.ndarray(m + 1)
n = np.zeros(m + 1, int)
for i in range(m + 1):
mask = d == i
p[i] = np.mean(input[mask])
n[i] = np.sum(mask)
return p, n
d = distance((10, 20))
center = (4, 5)
scale = 0.4
bin = 2
x, y, n = profile(d, center=center, scale=scale, bin=bin, histogram=True)
y_ref, n_ref = profile_slow(d, center=center, scale=scale, bin=bin)
assert_same(y, y_ref[0:y.size])
assert_same(n, n_ref[0:n.size])