def set_constraint(self, parameter, constraint_string=None,
constraint_dict=None):
r"""Set the constraint on a parameter/observable by specifying a string
or a dictionary. If several constraints (e.g. several types of
uncertainty) are given, the total constraint will be the convolution
of the individual distributions. Existing constraints will be removed.
Arguments:
- parameter: parameter string (or tuple)
- constraint_string: string specifying the constraint that can be e.g.
of the form `'1.55(3)(1)'` or `'4.0±0.1'`.
- constraint_dict: dictionary or list of several dictionaries of the
form `{'distribution': 'distribution_name', 'arg1': val1, ...}`, where
'distribution_name' is a string name associated to each probability
distribution (see `flavio.statistics.probability.class_from_string`)
and `'arg1'`, `val1` are argument/value pairs of the arguments of
the distribution class's constructor (e.g.`central_value`,
`standard_deviation` for a normal distribution).
`constraint_string` and `constraint_dict` must not be present
simultaneously.
"""
if constraint_string is not None and constraint_dict is not None:
raise ValueError("constraint_string and constraint_dict cannot"
" be used at the same time.")
if constraint_string is not None:
pds = constraints_from_string(constraint_string)
elif constraint_dict is not None:
pds = dict2dist(constraint_dict)
else:
raise TypeError("Either constraint_string or constraint_dict have"
" to be specified.")
combined_pd = convolve_distributions(pds)
self.add_constraint([parameter], combined_pd)
def set_constraint(self, parameter, constraint_string):
"""Set the constraints on a parameter/observable by specifying a string
that can be e.g. of the form 1.55(3)(1) or 4.0±0.1. Existing
constraints will be removed."""
pds = constraints_from_string(constraint_string)
combined_pd = convolve_distributions(pds)
self.add_constraint([parameter], combined_pd)
def set_constraint(self, parameter, constraint_string=None,
constraint_dict=None):
r"""Set the constraint on a parameter/observable by specifying a string
or a dictionary. If several constraints (e.g. several types of
uncertainty) are given, the total constraint will be the convolution
of the individual distributions. Existing constraints will be removed.
Arguments:
- parameter: parameter string (or tuple)
- constraint_string: string specifying the constraint that can be e.g.
of the form `'1.55(3)(1)'` or `'4.0±0.1'`.
- constraint_dict: dictionary or list of several dictionaries of the
form `{'distribution': 'distribution_name', 'arg1': val1, ...}`, where
'distribution_name' is a string name associated to each probability
distribution (see `flavio.statistics.probability.class_from_string`)
and `'arg1'`, `val1` are argument/value pairs of the arguments of
the distribution class's constructor (e.g.`central_value`,
`standard_deviation` for a normal distribution).
`constraint_string` and `constraint_dict` must not be present
simultaneously.
"""
if constraint_string is not None and constraint_dict is not None:
raise ValueError("constraint_string and constraint_dict cannot"
" be used at the same time.")
if constraint_string is not None:
pds = constraints_from_string(constraint_string)
elif constraint_dict is not None:
pds = dict2dist(constraint_dict)
else:
raise TypeError("Either constraint_string or constraint_dict have"
" to be specified.")
combined_pd = convolve_distributions(pds)
self.add_constraint([parameter], combined_pd)
def set_constraint(self, parameter, constraint_string):
"""Set the constraints on a parameter/observable by specifying a string
that can be e.g. of the form 1.55(3)(1) or 4.0±0.1. Existing
constraints will be removed."""
pds = constraints_from_string(constraint_string)
combined_pd = combine_distributions(pds)
self.add_constraint([parameter], combined_pd)
def _load(obj):
"""Read measurements from a YAML stream or file."""
measurements = yaml.load(obj)
for m_name, m_data in measurements.items():
m = Measurement(m_name)
for arg in ['inspire', 'hepdata', 'experiment', 'url']:
if arg in m_data:
setattr(m, arg, m_data[arg])
if 'correlation' not in m_data:
if isinstance(m_data['values'], list):
for value_dict in m_data['values']:
# if "value" is a list, it contains the values of observable
# arguments (like q^2)
args = Observable.get_instance(value_dict['name']).arguments
args_num = [value_dict[a] for a in args]
constraints = constraints_from_string(value_dict['value'])
error_dict = errors_from_string(value_dict['value'])
central_value = error_dict['central_value']
squared_error = 0.
for sym_err in error_dict['symmetric_errors']:
squared_error += sym_err**2
for asym_err in error_dict['asymmetric_errors']:
squared_error += asym_err[0]*asym_err[1]
args_num.insert(0, value_dict['name'])
obs_tuple = tuple(args_num)
m.add_constraint([obs_tuple], probability.NormalDistribution(central_value, sqrt(squared_error)))
else: # otherwise, 'values' is a dict just containing name: constraint_string
for obs, value in m_data['values'].items():
constraints = constraints_from_string(value)
for constraint in constraints:
m.add_constraint([obs], constraint)
else:
observables = []
central_values = []
errors = []
if isinstance(m_data['values'], list):
for value_dict in m_data['values']:
# if "value" is a list, it contains the values of observable
# arguments (like q^2)
args = Observable.get_instance(value_dict['name']).arguments
args_num = [value_dict[a] for a in args]
error_dict = errors_from_string(value_dict['value'])
args_num.insert(0, value_dict['name'])
obs_tuple = tuple(args_num)
observables.append(obs_tuple)
central_values.append(error_dict['central_value'])
squared_error = 0.
for sym_err in error_dict['symmetric_errors']:
squared_error += sym_err**2
for asym_err in error_dict['asymmetric_errors']:
squared_error += asym_err[0]*asym_err[1]
errors.append(sqrt(squared_error))
else: # otherwise, 'values' is a dict just containing name: constraint_string
for obs, value in m_data['values'].items():
observables.append(obs)
error_dict = errors_from_string(value)
central_values.append(error_dict['central_value'])
squared_error = 0.
for sym_err in error_dict['symmetric_errors']:
squared_error += sym_err**2
for asym_err in error_dict['asymmetric_errors']:
squared_error += asym_err[0]*asym_err[1]
errors.append(sqrt(squared_error))
correlation = _fix_correlation_matrix(m_data['correlation'], len(observables))
covariance = np.outer(np.asarray(errors), np.asarray(errors))*correlation
if not np.all(np.linalg.eigvals(covariance) > 0):
# if the covariance matrix is not positive definite, try a dirty trick:
# multiply all the correlations by 0.99.
n_dim = len(correlation)
correlation = (correlation - np.eye(n_dim))*0.99 + np.eye(n_dim)
covariance = np.outer(np.asarray(errors), np.asarray(errors))*correlation
# if it still isn't positive definite, give up.
assert np.all(np.linalg.eigvals(covariance) > 0), "The covariance matrix is not positive definite!" + str(covariance)
m.add_constraint(observables, probability.MultivariateNormalDistribution(central_values, covariance))
