prfmodel.regressors.RegressorsList

class prfmodel.regressors.RegressorsList(regressors: list[prfmodel.regressors.base.BaseRegressors])

Composite regressor model that sums the predictions of multiple regressor models.

Used internally by canonical models to support passing a list of regressor models as the regressors_model argument. The parameter names of all child regressor models are aggregated (preserving insertion order, removing duplicates).

At call time, the supplied design is a single pandas.DataFrame that is passed to every child, each of which slices the columns it needs by name.

Parameters:

regressors (list of BaseRegressors) – Non-empty list of regressor model instances.

Raises:

  • ValueError – If regressors is empty.

  • TypeError – If any element is not a BaseRegressors instance.

Examples

>>> import numpy as np
>>> import pandas as pd
>>> from prfmodel.regressors import AdditiveRegressors, RegressorsList
>>> a = AdditiveRegressors(names=["x"])
>>> b = AdditiveRegressors(names=["y"])
>>> regressors_model = RegressorsList([a, b])
>>> regressors_model.parameter_names
['beta_x', 'beta_y']
>>> params = pd.DataFrame({"beta_x": [1.0], "beta_y": [1.0]})
>>> design = pd.DataFrame({"x": np.ones(5), "y": np.ones(5) * 2.0})
>>> resp = regressors_model(design, params)
>>> print(resp.shape)
(1, 5)
__call__(regressors: pandas.DataFrame, parameters: pandas.DataFrame, dtype: str | None = None) → prfmodel.typing.Tensor

Compute the sum of predictions from all child regressor models.

Parameters:

  • regressors (pandas.DataFrame) – A single design data frame whose columns cover every child’s required regressor names. It is passed to each child, which slices the columns it needs by name; extra columns are ignored.

  • parameters (pandas.DataFrame) – Dataframe with columns containing different model parameters and rows containing parameter values for different units.

  • dtype (str, optional) – The dtype of the prediction result. If None (the default), uses the dtype from prfmodel.utils.get_dtype().

Returns:

The predicted model response with shape (num_units, num_frames) and dtype dtype.

Return type:

Tensor

property parameter_names: list[str]

Names of parameters used by the model, aggregated from all child regressor models.