Modifying default settings

As a user, you can easily

• change the default values of parameters and their uncertainties,
• choose between different parametrizations of various effects.

Understanding parameters & constraints

In flavio, all numerical quantities a theory prediction for an observable depends on are called parameters. Every parameter is typically subject to a constraint, i.e. a numerical value with an uncertainty that can stem from an experimental measurement or theoretical considerations. The simplest example is a mass: the parameter m_t corresponds to the top quark pole mass and by default is subject to the constraint $173.21\pm0.87$ GeV. A constraint can also apply to several parameters, to account for correlations. See a list of all parameters defined in the code. Note that all theory uncertainties are derived from uncertainties of these parameters. This means that also higher order effects have to parametrized by parameters.

When loading flavio, a set of default parameter constraints is loaded. All masses and widths are taken from the PDG; for some explanation of other parameter choices see here.

The default parameter constraints are taken from the following data files:

• for uncorrelated parameters, parameters_uncorrelated.yml, view on Github
• for uncorrelated parameters, parameters_correlated.yml, view on Github

Changing default values

There are two options.

In an interactive session or at the beginning of a script, you can directly overwrite the default parameter constraints by running, for instance,

flavio.default_parameters.set_constraint('m_t', '171.5(1.0)')

The first argument is always the name of an existing parameter, while the second argument is the constraint in the form of a string. Instead of the well-known bracket notation where the uncertainty refers to the last digit (e.g. '171.50(3)' for a ridiculously small error), you can also use a $\pm$ notation, e.g. '171.5 ± 1.0'. If you just specify a number, it will be treated as exact, i.e. without uncertainty.

The second option is to copy the YAML files linked above, modify them at will, and loading them back with

flavio.parameters.read_file_values('my_parameters.yml', flavio.default_parameters)

You can also just copy a subset of the default parameter file and only the values contained will be overwritten.

Changing the configuration

Some configuration options, in particular the default choices for parametrizations of various physical quantities like hadronic form factors, can be modified by modifying the configuration dictionary. The default options are read from the data file config.yml (view on Github).

There are two ways to modify them.

Either you directly change the dictionary flavio.config by running, e.g.,

flavio.config['implementation']['CKM matrix'] = 'Wolfenstein'

or you copy the YAML file linked above (or a subset of it), modify it at will, and load it back with

import yaml

with open('my_config.yml', 'r') as f:
  flavio.config.update(yaml.load(f))