Models

We use astropy.modeling.Model as model base. Similarly, all model instances can be combined as a astropy.modeling.CompoundModel instance using the + operator.

Implemented models

• Constant

• Gaussian

• DoublePowerLaw

• Beta

• ConeDoublePowerLaw

User model

For models that have not been implemented, it is easy to be defined using custom_model() decorator. Here is a simple example to create an unprojected power law model yourself.

import sbfit.model

@sbfit.model.custom_model
def PowerLaw(x, norm=1, alpha=1):
    return norm * x ** -alpha

my_pl = PowerLaw()

Note

The first argument of a user model should be x. Model parameters are given from the second argument with default values. To follow the convention, the first parameter should always be the normalization norm.