Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add support for using toymodel in telescope frame #2349

Merged
merged 5 commits into from
Jun 20, 2023
Merged

Add support for using toymodel in telescope frame #2349

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
c850540
Add support for using toymodel in telescope frame
maxnoe Jun 13, 2023
e18d5f4
Update docstrings
maxnoe Jun 19, 2023
195fdb8
Remove unneeded funciton
maxnoe Jun 19, 2023
98bf240
Only use length in docstring
maxnoe Jun 19, 2023
974f7bc
Add docstring
maxnoe Jun 19, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
120 changes: 83 additions & 37 deletions ctapipe/image/tests/test_toy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,11 +6,13 @@
from pytest import approx
from scipy.stats import norm, poisson, skewnorm

from ctapipe.coordinates.telescope_frame import TelescopeFrame
from ctapipe.image.toymodel import WaveformModel, obtain_time_image


@pytest.mark.parametrize("frame", ["telescope", "camera"])
@pytest.mark.parametrize("seed", [None, 0])
def test_intensity(seed, monkeypatch, prod5_lst):
def test_intensity(seed, frame, monkeypatch, prod5_lst):
"""
Test generation of the toymodel roughly follows the given intensity.

Expand All @@ -19,13 +21,19 @@ def test_intensity(seed, monkeypatch, prod5_lst):
"""
from ctapipe.image import toymodel

geom = prod5_lst.camera.geometry
if frame == "camera":
geom = prod5_lst.camera.geometry
unit = u.m
else:
geom = prod5_lst.camera.geometry.transform_to(TelescopeFrame())
unit = u.deg

x, y = u.Quantity([0.2, 0.3], unit)
width = 0.05 * unit
length = 0.15 * unit

x, y = u.Quantity([0.2, 0.3], u.m)
width = 0.05 * u.m
length = 0.15 * u.m
intensity = 200
psi = "30d"
psi = 30 * u.deg

# make sure we set a fixed seed for this test even when testing the
# API without giving the rng
Expand All @@ -43,33 +51,43 @@ def test_intensity(seed, monkeypatch, prod5_lst):
)

# test if signal reproduces given cog values
assert np.average(geom.pix_x.to_value(u.m), weights=signal) == approx(0.2, rel=0.15)
assert np.average(geom.pix_y.to_value(u.m), weights=signal) == approx(0.3, rel=0.15)
assert np.average(geom.pix_x.to_value(unit), weights=signal) == approx(
0.2, rel=0.15
)
assert np.average(geom.pix_y.to_value(unit), weights=signal) == approx(
0.3, rel=0.15
)

# test if signal reproduces given width/length values
cov = np.cov(geom.pix_x.value, geom.pix_y.value, aweights=signal)
eigvals, _ = np.linalg.eigh(cov)

assert np.sqrt(eigvals[0]) == approx(width.to_value(u.m), rel=0.15)
assert np.sqrt(eigvals[1]) == approx(length.to_value(u.m), rel=0.15)
assert np.sqrt(eigvals[0]) == approx(width.to_value(unit), rel=0.15)
assert np.sqrt(eigvals[1]) == approx(length.to_value(unit), rel=0.15)

# test if total intensity is inside in 99 percent confidence interval
assert poisson(intensity).ppf(0.05) <= signal.sum() <= poisson(intensity).ppf(0.95)


def test_skewed(prod5_lst):
@pytest.mark.parametrize("frame", ["telescope", "camera"])
def test_skewed(frame, prod5_lst):
from ctapipe.image.toymodel import SkewedGaussian

# test if the parameters we calculated for the skew normal
# distribution produce the correct moments
rng = np.random.default_rng(0)
geom = prod5_lst.camera.geometry
if frame == "camera":
geom = prod5_lst.camera.geometry
unit = u.m
else:
geom = prod5_lst.camera.geometry.transform_to(TelescopeFrame())
unit = u.deg

x, y = u.Quantity([0.2, 0.3], u.m)
width = 0.05 * u.m
length = 0.15 * u.m
x, y = u.Quantity([0.2, 0.3], unit)
width = 0.05 * unit
length = 0.15 * unit
intensity = 50
psi = "30d"
psi = 30 * u.deg
skewness = 0.3

model = SkewedGaussian(
Expand All @@ -81,20 +99,26 @@ def test_skewed(prod5_lst):
mean, var, skew = skewnorm(a=a, loc=loc, scale=scale).stats(moments="mvs")

assert np.isclose(mean, 0)
assert np.isclose(var, length.to_value(u.m) ** 2)
assert np.isclose(var, length.to_value(unit) ** 2)
assert np.isclose(skew, skewness)


def test_compare(prod5_lst):
@pytest.mark.parametrize("frame", ["telescope", "camera"])
def test_compare(frame, prod5_lst):
from ctapipe.image.toymodel import Gaussian, SkewedGaussian

geom = prod5_lst.camera.geometry
if frame == "camera":
geom = prod5_lst.camera.geometry
unit = u.m
else:
geom = prod5_lst.camera.geometry.transform_to(TelescopeFrame())
unit = u.deg

x, y = u.Quantity([0.2, 0.3], u.m)
width = 0.05 * u.m
length = 0.15 * u.m
x, y = u.Quantity([0.2, 0.3], unit)
width = 0.05 * unit
length = 0.15 * unit
intensity = 50
psi = "30d"
psi = 30 * u.deg

skewed = SkewedGaussian(x=x, y=y, width=width, length=length, psi=psi, skewness=0)
normal = Gaussian(x=x, y=y, width=width, length=length, psi=psi)
Expand All @@ -105,14 +129,29 @@ def test_compare(prod5_lst):
assert np.isclose(signal_skewed, signal_normal).all()


def test_obtain_time_image(prod5_sst):
@pytest.mark.parametrize("frame", ["telescope", "camera"])
def test_obtain_time_image(frame, prod5_sst):
geom = prod5_sst.camera.geometry

centroid_x = u.Quantity(0.05, u.m)
centroid_y = u.Quantity(0.05, u.m)
if frame == "camera":
geom = prod5_sst.camera.geometry
unit = u.m
scale = 1.0
else:
geom_cam_frame = prod5_sst.camera.geometry
geom = geom_cam_frame.transform_to(TelescopeFrame())
unit = u.deg
# further down we test the std deviation, but that scales
# with the size of our shower in the camera
r_tel_frame = geom.guess_radius().to_value(u.deg)
r_cam_frame = geom_cam_frame.guess_radius().to_value(u.m)
scale = r_tel_frame / r_cam_frame

centroid_x = u.Quantity(0.05, unit)
centroid_y = u.Quantity(0.05, unit)
psi = u.Quantity(70, u.deg)

time_gradient = u.Quantity(0, u.ns / u.m)
time_gradient = u.Quantity(0, u.ns / unit)
time_intercept = u.Quantity(0, u.ns)
time = obtain_time_image(
geom.pix_x,
Expand All @@ -125,7 +164,7 @@ def test_obtain_time_image(prod5_sst):
)
np.testing.assert_allclose(time, 0)

time_gradient = u.Quantity(0, u.ns / u.m)
time_gradient = u.Quantity(0, u.ns / unit)
time_intercept = u.Quantity(40, u.ns)
time = obtain_time_image(
geom.pix_x,
Expand All @@ -138,7 +177,7 @@ def test_obtain_time_image(prod5_sst):
)
np.testing.assert_allclose(time, 40)

time_gradient = u.Quantity(20, u.ns / u.m)
time_gradient = u.Quantity(20 / scale, u.ns / unit)
time_intercept = u.Quantity(40, u.ns)
time = obtain_time_image(
geom.pix_x,
Expand All @@ -151,7 +190,7 @@ def test_obtain_time_image(prod5_sst):
)
np.testing.assert_allclose(time.std(), 1.710435, rtol=0.1)

time_gradient = u.Quantity(20, u.ns / u.m)
time_gradient = u.Quantity(20, u.ns / unit)
time_intercept = u.Quantity(40, u.ns)
time = obtain_time_image(
centroid_x,
Expand All @@ -165,13 +204,20 @@ def test_obtain_time_image(prod5_sst):
np.testing.assert_allclose(time, 40)


def test_waveform_model(prod5_sst):
@pytest.mark.parametrize("frame", ["telescope", "camera"])
def test_waveform_model(frame, prod5_sst):
from ctapipe.image.toymodel import Gaussian

prod5_sst = deepcopy(prod5_sst)
geom = prod5_sst.camera.geometry
readout = prod5_sst.camera.readout

if frame == "camera":
geom = prod5_sst.camera.geometry
unit = u.m
else:
geom = prod5_sst.camera.geometry.transform_to(TelescopeFrame())
unit = u.deg

ref_duration = 67
n_ref_samples = 100
pulse_sigma = 3
Expand All @@ -184,12 +230,12 @@ def test_waveform_model(prod5_sst):
)
readout.sampling_rate = u.Quantity(2, u.GHz)

centroid_x = u.Quantity(0.05, u.m)
centroid_y = u.Quantity(0.05, u.m)
length = u.Quantity(0.03, u.m)
width = u.Quantity(0.008, u.m)
centroid_x = u.Quantity(0.05, unit)
centroid_y = u.Quantity(0.05, unit)
length = u.Quantity(0.03, unit)
width = u.Quantity(0.008, unit)
psi = u.Quantity(70, u.deg)
time_gradient = u.Quantity(50, u.ns / u.m)
time_gradient = u.Quantity(50, u.ns / unit)
time_intercept = u.Quantity(20, u.ns)

_, charge, _ = Gaussian(
Expand Down
Loading