skneuromsi.ndcollection.bias_acc module

Implementation of bias analysis in multisensory integration.

The NDResultBiasAcc class provides an accessor, NDResultBiasAcc, for analyzing biases in the context of multisensory integration. It offers methods to calculate biases and mean biases based on specified influence and changing parameters.

class skneuromsi.ndcollection.bias_acc.NDResultCollectionBiasAcc(ndcollection, tqdm_cls)[source]

Bases: AccessorABC

Accessor for calculating biases in an NDResultCollection.

Bias analysis in multisensory integration refers to the examination and identification of potential biases that may influence the way different sensory modalities are combined or integrated in the human brain.

This accessor provides methods for calculating biases based on various parameters.

Parameters:

  • ndcollection (NDResultCollection) – The NDResultCollection for which to calculate biases.

  • tqdm_cls (tqdm.tqdm, optional) – The tqdm class to use. Defaults to None.

bias = <methodtools._LruCacheWire object>[source]
bias_mean(influence_parameter, *, changing_parameter=None, dim=None, mode=None, quiet=False)[source]

Calculate the mean biases in multisensory integration analysis.

This method calculates the mean biases in the context of multisensory integration analysis. Mean biases represent the average deviation of a specific parameter’s influence on the integration process across different iterations. The analysis considers the relationship between the changing parameter and the influence parameter in each iteration.

Parameters:

  • influence_parameter (str) – The parameter being influenced by the cross-modal biases.

  • changing_parameter (str or None, optional) – The parameter that changes across iterations. If None, the function automatically selects it.

  • dim (str or None, optional) – The dimension for calculating mean biases. If None, it defaults to ‘time’.

  • mode (str or None, optional) – The mode for which to calculate mean biases. If None, it defaults to the mode with maximum variance.

  • quiet (bool, optional) – If True, suppresses the tqdm progress bar. Defaults to False.

Returns:

A DataFrame containing the mean biases in multisensory integration analysis. The columns represent the changing and influence parameters, and each row corresponds to a specific disparity in the relationship between these parameters across iterations.

Return type:

pandas.DataFrame

Raises:

ValueError – If the specified parameters are invalid or the influence parameter is not fixed across iterations.