Skip to content

Documentation for BaseModelWrapper

The PHOTONAI interface for implementing custom pipeline elements.

PHOTONAI works on top of the scikit-learn object API, see documentation.

Your class should overwrite the following definitions:

  • fit(data): learn or adjust to the data.

If it is an estimator, which means it has the ability to learn,

  • it should implement predict(data): using the learned model to generate prediction,
  • should inherit sklearn.base.BaseEstimator (see here),
  • inherits get_params and set_params.

If it is an transformer, which means it preprocesses or prepares the data,

  • it should implement transform(data): applying the logic to the data to transform it,
  • should inherit from sklearn.base.TransformerMixin (see here),
  • inherits fit_transform as a concatenation of both fit and transform,
  • should inherit sklearn.base.BaseEstimator (see here)
  • inherits get_params and set_params.

Prepare for hyperparameter optimization

PHOTONAI expects a definition for all parameters you want to optimize in the hyperparameter search in the constructor stub, and to be addressable with the same name as class variable. In this way you can define any parameter and it is automatically prepared for the hyperparameter search process.

See the scikit-learn object API documentation for more in depth information about the interface.

fit(self, data, targets=None)

Adjust the underlying model or method to the data.

Parameters:

Name Type Description Default
data ndarray

The input samples of shape [n_samples, n_original_features].

required
targets ndarray

The input targets of shape [n_samples, 1].

None

Returns:

Type Description

IMPORTANT, must return self!

Source code in photonai/modelwrapper/base_model_wrapper.py
def fit(self, data: np.ndarray, targets: np.ndarray = None):
    """
    Adjust the underlying model or method to the data.

    Parameters:
        data:
            The input samples of shape [n_samples, n_original_features].

        targets:
            The input targets of shape [n_samples, 1].

    Returns:
        IMPORTANT, must return self!

    """

get_params(self, deep=True)

Get the models parameters. Automatically implemented when inheriting from sklearn.base.BaseEstimator

Parameters:

Name Type Description Default
deep bool

If True, will return the parameters for this estimator and contained subobjects that are estimators.

True

Returns:

Type Description
dict

Parameter names mapped to their values.

Source code in photonai/modelwrapper/base_model_wrapper.py
def get_params(self, deep: bool = True) -> dict:
    """
    Get the models parameters.
    Automatically implemented when inheriting from sklearn.base.BaseEstimator

    Parameters:
        deep:
            If True, will return the parameters for this estimator and contained subobjects that are estimators.

    Returns:
        Parameter names mapped to their values.
    """
    return super(BaseModelWrapper, self).get_params(deep=deep)

predict(self, data)

Use the learned model to make predictions.

Parameters:

Name Type Description Default
data ndarray

The input samples of shape [n_samples, n_original_features].

required
Source code in photonai/modelwrapper/base_model_wrapper.py
def predict(self, data: np.ndarray):
    """
    Use the learned model to make predictions.

    Parameters:
        data:
            The input samples of shape [n_samples, n_original_features].

    """

set_params(self, **kwargs)

Takes the given dictionary, with the keys being the variable name, and sets the object's parameters to the given values. Automatically implemented when inheriting from sklearn.base.BaseEstimator.

Parameters:

Name Type Description Default
**kwargs

Estimator parameters.

{}
Source code in photonai/modelwrapper/base_model_wrapper.py
def set_params(self, **kwargs):
    """
    Takes the given dictionary, with the keys being the variable name,
    and sets the object's parameters to the given values.
    Automatically implemented when inheriting from sklearn.base.BaseEstimator.

    Parameters:
        **kwargs: Estimator parameters.

    """
    super(BaseModelWrapper, self).set_params(**kwargs)

transform(self, data, targets=None)

Apply the method's logic to the data.

Parameters:

Name Type Description Default
data ndarray

The input samples of shape [n_samples, n_original_features].

required
targets ndarray

The input targets of shape [n_samples, 1]. Not necessary.

None
Source code in photonai/modelwrapper/base_model_wrapper.py
def transform(self, data: np.ndarray, targets: np.ndarray = None):
    """
    Apply the method's logic to the data.

    Parameters:
        data:
            The input samples of shape [n_samples, n_original_features].

        targets:
            The input targets of shape [n_samples, 1]. Not necessary.

    """