def test_goodness_of_fit():
# Let's generate some data with y = Powerlaw(x)
gen_function = Powerlaw()
# Generate a dataset using the power law, and a
# constant 30% error
x = np.logspace(0, 2, 50)
xyl_generator = XYLike.from_function(
"sim_data", function=gen_function, x=x, yerr=0.3 * gen_function(x)
)
y = xyl_generator.y
y_err = xyl_generator.yerr
fit_function = Powerlaw()
xyl = XYLike("data", x, y, y_err)
parameters, like_values = xyl.fit(fit_function)
gof, all_results, all_like_values = xyl.goodness_of_fit()
# Compute the number of degrees of freedom
n_dof = len(xyl.x) - len(fit_function.free_parameters)
# Get the observed value for chi2
obs_chi2 = 2 * like_values["-log(likelihood)"]["data"]
theoretical_gof = scipy.stats.chi2(n_dof).sf(obs_chi2)
assert np.isclose(theoretical_gof, gof["total"], rtol=0.1)
def test_XYLike_chi2():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
# Verify that the fit converged where it should have
assert np.allclose(res[0]['value'].values,[0.82896119, 40.20269202, 62.80359114, 5.04080011, 0.27286713], rtol=0.05)
# test not setting yerr
xy = XYLike("test", x, y)
assert np.all(xy.yerr == np.ones_like(y))
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
def test_XYLike_assign_to_source():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
xy.assign_to_source("pts1")
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
fitfun2 = Line() + Gaussian()
fitfun2.F_2 = 60.0
fitfun2.mu_2 = 4.5
pts1 = PointSource("pts1", ra=0.0, dec=0.0, spectral_shape=fitfun)
pts2 = PointSource("pts2", ra=2.5, dec=3.2, spectral_shape=fitfun2)
for parameter in fitfun2.parameters.values():
parameter.fix = True
model = Model(pts1, pts2)
data = DataList(xy)
jl = JointLikelihood(model, data)
_ = jl.fit()
predicted_parameters = np.array([0.82896119, 40.20269202, 62.80359114, 5.04080011, 0.27286713])
assert np.allclose([fitfun.a_1.value, fitfun.b_1.value, fitfun.F_2.value, fitfun.mu_2.value, fitfun.sigma_2.value],
predicted_parameters, rtol=0.05)
# Test that the likelihood does not change by changing the parameters of the other source
log_like_before = jl.minus_log_like_profile(*predicted_parameters)
fitfun2.F_2 = 120.0
log_like_after = jl.minus_log_like_profile(*predicted_parameters)
assert log_like_before == log_like_after
# Now test that if we do not assign a source, then the log likelihood value will change
xy.assign_to_source(None)
# Test that the likelihood this time changes by changing the parameters of the other source
log_like_before = jl.minus_log_like_profile(*predicted_parameters)
fitfun2.F_2 = 60.0
log_like_after = jl.minus_log_like_profile(*predicted_parameters)
assert log_like_before != log_like_after
def test_XYLike_dataframe():
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# chi2 version
xy = XYLike("test", x, y, yerr)
df = xy.to_dataframe()
# read back in dataframe
new_xy = XYLike.from_dataframe("df", df)
assert not xy.is_poisson
# poisson version
xy = XYLike("test", x, y, poisson_data=True)
df = xy.to_dataframe()
# read back in dataframe
new_xy = XYLike.from_dataframe("df", df, poisson=True)
assert xy.is_poisson
def test_XYLike_poisson():
# Now Poisson case
y = np.array(poiss_sig)
xy = XYLike("test", x, y, poisson_data=True)
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.F_2.bounds = (0, 200.0)
fitfun.mu_2 = 5.0
fitfun.a_1.bounds = (0.1, 5.0)
fitfun.b_1.bounds = (0.1, 100.0)
res = xy.fit(fitfun)
# Verify that the fit converged where it should have
#print res[0]['value']
assert np.allclose(res[0]['value'], [0.783748,40.344599 , 71.560055, 4.989727 , 0.330570 ], rtol=0.05)
def test_XYLike_dataframe():
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# chi2 version
xy = XYLike("test", x, y, yerr)
df = xy.to_dataframe()
# read back in dataframe
new_xy = XYLike.from_dataframe('df', df)
assert not xy.is_poisson
# poisson version
xy = XYLike("test", x, y, poisson_data=True)
df = xy.to_dataframe()
# read back in dataframe
new_xy = XYLike.from_dataframe('df', df, poisson=True)
assert xy.is_poisson
def test_XYLike_chi2():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
# Verify that the fit converged where it should have
assert np.allclose(
res[0]["value"].values,
[0.82896119, 40.20269202, 62.80359114, 5.04080011, 0.27286713],
rtol=0.05,
)
# test not setting yerr
xy = XYLike("test", x, y)
assert np.all(xy.yerr == np.ones_like(y))
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
def test_XYLike_poisson():
# Now Poisson case
y = np.array(poiss_sig)
xy = XYLike("test", x, y, poisson_data=True)
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.F_2.bounds = (0, 200.0)
fitfun.mu_2 = 5.0
fitfun.a_1.bounds = (0.1, 5.0)
fitfun.b_1.bounds = (0.1, 100.0)
res = xy.fit(fitfun)
# Verify that the fit converged where it should have
# print res[0]['value']
assert np.allclose(res[0]["value"],
[0.783748, 40.344599, 71.560055, 4.989727, 0.330570],
rtol=0.05)
def test_XYLike_txt():
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# chi2 version
xy = XYLike("test", x, y, yerr)
fname = "test_txt.txt"
xy.to_txt(fname)
# read back in txt file
new_xy = XYLike.from_text_file("txt", fname)
assert not xy.is_poisson
# poisson version
xy = XYLike("test", x, y, poisson_data=True)
fname = "test_txt_poisson.txt"
xy.to_txt(fname)
# read back in txt file
new_xy = XYLike.from_text_file("txt", fname)
assert new_xy.is_poisson
# Remove files
os.remove("test_txt.txt")
os.remove("test_txt_poisson.txt")
def test_goodness_of_fit():
# Let's generate some data with y = Powerlaw(x)
gen_function = Powerlaw()
# Generate a dataset using the power law, and a
# constant 30% error
x = np.logspace(0, 2, 50)
xyl_generator = XYLike.from_function("sim_data", function=gen_function,
x=x,
yerr=0.3 * gen_function(x))
y = xyl_generator.y
y_err = xyl_generator.yerr
fit_function = Powerlaw()
xyl = XYLike("data", x, y, y_err)
parameters, like_values = xyl.fit(fit_function)
gof, all_results, all_like_values = xyl.goodness_of_fit()
# Compute the number of degrees of freedom
n_dof = len(xyl.x) - len(fit_function.free_parameters)
# Get the observed value for chi2
obs_chi2 = 2 * like_values['-log(likelihood)']['data']
theoretical_gof = scipy.stats.chi2(n_dof).sf(obs_chi2)
assert np.isclose(theoretical_gof, gof['total'], rtol=0.1)
def xy_model_and_datalist():
y = np.array(poiss_sig)
xy = XYLike("test", x, y, poisson_data=True)
fitfun = Line() + Gaussian()
fitfun.b_1.bounds = (-10, 10.0)
fitfun.a_1.bounds = (-100, 100.0)
fitfun.F_2 = 60.0
fitfun.F_2.bounds = (1e-3, 200.0)
fitfun.mu_2 = 5.0
fitfun.mu_2.bounds = (0.0, 100.0)
fitfun.sigma_2.bounds = (1e-3, 10.0)
model = Model(PointSource("fake", 0.0, 0.0, fitfun))
data = DataList(xy)
return model, data
def test_xy_plot():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
xy.plot()
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
xy.plot()
def test_XYLike_txt():
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# chi2 version
xy = XYLike("test", x, y, yerr)
fname = 'test_txt.txt'
xy.to_txt(fname)
# read back in txt file
new_xy = XYLike.from_text_file('txt', fname )
assert not xy.is_poisson
# poisson version
xy = XYLike("test", x, y, poisson_data=True)
fname = 'test_txt_poisson.txt'
xy.to_txt(fname)
# read back in txt file
new_xy = XYLike.from_text_file('txt', fname)
assert new_xy.is_poisson
# Remove files
os.remove('test_txt.txt')
os.remove('test_txt_poisson.txt')
def generate_one(K):
# Let's generate some data with y = Powerlaw(x)
gen_function = Powerlaw()
gen_function.K = K
# Generate a dataset using the power law, and a
# constant 30% error
x = np.logspace(0, 2, 50)
xyl_generator = XYLike.from_function("sim_data", function=gen_function,
x=x,
yerr=0.3 * gen_function(x))
y = xyl_generator.y
y_err = xyl_generator.yerr
# xyl = XYLike("data", x, y, y_err)
# xyl.plot(x_scale='log', y_scale='log')
return x, y, y_err
def test_xy_plot():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
xy.plot()
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
res = xy.fit(fitfun)
xy.plot()
def test_energy_time_fit():
# Let's generate our dataset of 4 spectra with a normalization that follows
# a powerlaw in time
def generate_one(K):
# Let's generate some data with y = Powerlaw(x)
gen_function = Powerlaw()
gen_function.K = K
# Generate a dataset using the power law, and a
# constant 30% error
x = np.logspace(0, 2, 50)
xyl_generator = XYLike.from_function("sim_data", function=gen_function,
x=x,
yerr=0.3 * gen_function(x))
y = xyl_generator.y
y_err = xyl_generator.yerr
# xyl = XYLike("data", x, y, y_err)
# xyl.plot(x_scale='log', y_scale='log')
return x, y, y_err
time_tags = np.array([1.0, 2.0, 5.0, 10.0])
# This is the power law that defines the normalization as a function of time
normalizations = 0.23 * time_tags ** (-1.2)
datasets = map(generate_one, normalizations)
# Now set up the fit and fit it
time = IndependentVariable("time", 1.0, u.s)
plugins = []
for i, dataset in enumerate(datasets):
x, y, y_err = dataset
xyl = XYLike("data%i" % i, x, y, y_err)
xyl.tag = (time, time_tags[i])
assert xyl.tag == (time, time_tags[i], None)
plugins.append(xyl)
data = DataList(*plugins)
spectrum = Powerlaw()
spectrum.K.bounds = (0.01, 1000.0)
src = PointSource("test", 0.0, 0.0, spectrum)
model = Model(src)
model.add_independent_variable(time)
time_po = Powerlaw()
time_po.K.bounds = (0.01, 1000)
time_po.K.value = 2.0
time_po.index = -1.5
model.link(spectrum.K, time, time_po)
jl = JointLikelihood(model, data)
jl.set_minimizer("minuit")
best_fit_parameters, likelihood_values = jl.fit()
# Make sure we are within 10% of the expected result
assert np.allclose(best_fit_parameters['value'].values, [0.25496115, -1.2282951 , -2.01508341], rtol=0.1)
def test_XYLike_assign_to_source():
# Get fake data with Gaussian noise
yerr = np.array(gauss_sigma)
y = np.array(gauss_signal)
# Fit
xy = XYLike("test", x, y, yerr)
xy.assign_to_source("pts1")
fitfun = Line() + Gaussian()
fitfun.F_2 = 60.0
fitfun.mu_2 = 4.5
fitfun2 = Line() + Gaussian()
fitfun2.F_2 = 60.0
fitfun2.mu_2 = 4.5
pts1 = PointSource("pts1", ra=0.0, dec=0.0, spectral_shape=fitfun)
pts2 = PointSource("pts2", ra=2.5, dec=3.2, spectral_shape=fitfun2)
for parameter in list(fitfun2.parameters.values()):
parameter.fix = True
model = Model(pts1, pts2)
data = DataList(xy)
jl = JointLikelihood(model, data)
_ = jl.fit()
predicted_parameters = np.array(
[0.82896119, 40.20269202, 62.80359114, 5.04080011, 0.27286713])
assert np.allclose(
[
fitfun.a_1.value,
fitfun.b_1.value,
fitfun.F_2.value,
fitfun.mu_2.value,
fitfun.sigma_2.value,
],
predicted_parameters,
rtol=0.05,
)
# Test that the likelihood does not change by changing the parameters of the other source
log_like_before = jl.minus_log_like_profile(*predicted_parameters)
fitfun2.F_2 = 120.0
log_like_after = jl.minus_log_like_profile(*predicted_parameters)
assert log_like_before == log_like_after
# Now test that if we do not assign a source, then the log likelihood value will change
xy.assign_to_source(None)
# Test that the likelihood this time changes by changing the parameters of the other source
log_like_before = jl.minus_log_like_profile(*predicted_parameters)
fitfun2.F_2 = 60.0
log_like_after = jl.minus_log_like_profile(*predicted_parameters)
assert log_like_before != log_like_after
def polyfit(x: Iterable[float], y: Iterable[float], grade: int, exposure: Iterable[float], bayes: bool = False) -> Tuple[Polynomial, float]:
"""
function to fit a polynomial to data.
not a member to allow parallel computation
:param x: the x coord of the data
:param y: teh y coord of the data
:param grade: the polynomical order or grade
:param expousure: the exposure of the interval
:param bayes: to do a bayesian fit or not
"""
# Check that we have enough counts to perform the fit, otherwise
# return a "zero polynomial"
log.debug(f"starting polyfit with grade {grade} ")
if threeML_config.time_series.default_fit_method is not None:
bayes = threeML_config.time_series.default_fit_method
log.debug("using a default poly fit method")
nan_mask = np.isnan(y)
y = y[~nan_mask]
x = x[~nan_mask]
exposure = exposure[~nan_mask]
non_zero_mask = y > 0
n_non_zero = non_zero_mask.sum()
if n_non_zero == 0:
log.debug("no counts, return 0")
# No data, nothing to do!
return Polynomial([0.0]*(grade+1)), 0.0
# create 3ML plugins and fit them with 3ML!
# should eventuallly allow better config
# seelct the model based on the grade
shape = _grade_model_lookup[grade]()
ps = PointSource("_dummy", 0, 0, spectral_shape=shape)
model = Model(ps)
avg = np.mean(y/exposure)
log.debug(f"starting polyfit with avg norm {avg}")
with silence_console_log():
xy = XYLike("series", x=x, y=y, exposure=exposure,
poisson_data=True, quiet=True)
if not bayes:
# make sure the model is positive
for i, (k, v) in enumerate(model.free_parameters.items()):
if i == 0:
v.bounds = (0, None)
v.value = avg
else:
v.value = 0.0
# we actually use a line here
# because a constant is returns a
# single number
if grade == 0:
shape.b = 0
shape.b.fix = True
jl: JointLikelihood = JointLikelihood(model, DataList(xy))
jl.set_minimizer("minuit")
# if the fit falis, retry and then just accept
try:
jl.fit(quiet=True)
except(FitFailed, BadCovariance, AllFitFailed, CannotComputeCovariance):
log.debug("1st fit failed")
try:
jl.fit(quiet=True)
except(FitFailed, BadCovariance, AllFitFailed, CannotComputeCovariance):
log.debug("all MLE fits failed")
pass
coeff = [v.value for _, v in model.free_parameters.items()]
log.debug(f"got coeff: {coeff}")
final_polynomial = Polynomial(coeff)
try:
final_polynomial.set_covariace_matrix(
jl.results.covariance_matrix)
except:
log.exception(f"Fit failed in channel")
raise FitFailed()
min_log_likelihood = xy.get_log_like()
else:
# set smart priors
for i, (k, v) in enumerate(model.free_parameters.items()):
if i == 0:
v.bounds = (0, None)
v.prior = Log_normal(
mu=np.log(avg), sigma=np.max([np.log(avg/2), 1]))
v.value = 1
else:
v.prior = Gaussian(mu=0, sigma=2)
v.value = 1e-2
# we actually use a line here
# because a constant is returns a
# single number
if grade == 0:
shape.b = 0
shape.b.fix = True
ba: BayesianAnalysis = BayesianAnalysis(model, DataList(xy))
ba.set_sampler("emcee")
ba.sampler.setup(n_iterations=500, n_burn_in=200, n_walkers=20)
ba.sample(quiet=True)
ba.restore_median_fit()
coeff = [v.value for _, v in model.free_parameters.items()]
log.debug(f"got coeff: {coeff}")
final_polynomial = Polynomial(coeff)
final_polynomial.set_covariace_matrix(
ba.results.estimate_covariance_matrix())
min_log_likelihood = xy.get_log_like()
log.debug(f"-min loglike: {-min_log_likelihood}")
return final_polynomial, -min_log_likelihood
def test_energy_time_fit():
# Let's generate our dataset of 4 spectra with a normalization that follows
# a powerlaw in time
def generate_one(K):
# Let's generate some data with y = Powerlaw(x)
gen_function = Powerlaw()
gen_function.K = K
# Generate a dataset using the power law, and a
# constant 30% error
x = np.logspace(0, 2, 50)
xyl_generator = XYLike.from_function("sim_data",
function=gen_function,
x=x,
yerr=0.3 * gen_function(x))
y = xyl_generator.y
y_err = xyl_generator.yerr
# xyl = XYLike("data", x, y, y_err)
# xyl.plot(x_scale='log', y_scale='log')
return x, y, y_err
time_tags = np.array([1.0, 2.0, 5.0, 10.0])
# This is the power law that defines the normalization as a function of time
normalizations = 0.23 * time_tags**(-1.2)
datasets = list(map(generate_one, normalizations))
# Now set up the fit and fit it
time = IndependentVariable("time", 1.0, u.s)
plugins = []
for i, dataset in enumerate(datasets):
x, y, y_err = dataset
xyl = XYLike("data%i" % i, x, y, y_err)
xyl.tag = (time, time_tags[i])
assert xyl.tag == (time, time_tags[i], None)
plugins.append(xyl)
data = DataList(*plugins)
spectrum = Powerlaw()
spectrum.K.bounds = (0.01, 1000.0)
src = PointSource("test", 0.0, 0.0, spectrum)
model = Model(src)
model.add_independent_variable(time)
time_po = Powerlaw()
time_po.K.bounds = (0.01, 1000)
time_po.K.value = 2.0
time_po.index = -1.5
model.link(spectrum.K, time, time_po)
jl = JointLikelihood(model, data)
jl.set_minimizer("minuit")
best_fit_parameters, likelihood_values = jl.fit()
# Make sure we are within 10% of the expected result
assert np.allclose(
best_fit_parameters["value"].values,
[0.25496115, -1.2282951, -2.01508341],
rtol=0.1,
)
