modalities.fmri.experimental_paradigm

Module: modalities.fmri.experimental_paradigm

Inheritance diagram for nipy.modalities.fmri.experimental_paradigm:

This module implements an object to deal with experimental paradigms. In fMRI data analysis, there are two main types of experimental paradigms: block and event-related paradigms. They correspond to 2 classes EventRelatedParadigm and BlockParadigm. Both are implemented here, together with functions to write paradigms to csv files.

Notes

Although the Paradigm object have no notion of session or acquisitions (they are assumed to correspond to a sequential acquisition, called ‘session’ in SPM jargon), the .csv file used to represent paradigm may be multi-session, so it is assumed that the first column of a file yielding a paradigm is in fact a session index

Author: Bertrand Thirion, 2009-2011

Classes

BlockParadigm

class nipy.modalities.fmri.experimental_paradigm.BlockParadigm(con_id=None, onset=None, duration=None, amplitude=None)

Bases: Paradigm

Class to handle block paradigms

__init__(con_id=None, onset=None, duration=None, amplitude=None)

Parameters:

con_id: array of shape (n_events), type = string, optional

id of the events (name of the experimental condition)

onset: array of shape (n_events), type = float, optional

onset time (in s.) of the events

amplitude: array of shape (n_events), type = float, optional,

amplitude of the events (if applicable)

write_to_csv(csv_file, session='0')

Write the paradigm to a csv file

Parameters:

csv_file: string, path of the csv file

session: string, optional, session identifier

EventRelatedParadigm

class nipy.modalities.fmri.experimental_paradigm.EventRelatedParadigm(con_id=None, onset=None, amplitude=None)

Bases: Paradigm

Class to handle event-related paradigms

__init__(con_id=None, onset=None, amplitude=None)

Parameters:

con_id: array of shape (n_events), type = string, optional

id of the events (name of the experimental condition)

onset: array of shape (n_events), type = float, optional

onset time (in s.) of the events

amplitude: array of shape (n_events), type = float, optional,

amplitude of the events (if applicable)

write_to_csv(csv_file, session='0')

Write the paradigm to a csv file

Parameters:

csv_file: string, path of the csv file

session: string, optional, session identifier

Paradigm

class nipy.modalities.fmri.experimental_paradigm.Paradigm(con_id=None, onset=None, amplitude=None)

Bases: object

Simple class to handle the experimental paradigm in one session

__init__(con_id=None, onset=None, amplitude=None)

Parameters:

con_id: array of shape (n_events), type = string, optional

identifier of the events

onset: array of shape (n_events), type = float, optional,

onset time (in s.) of the events

amplitude: array of shape (n_events), type = float, optional,

amplitude of the events (if applicable)

write_to_csv(csv_file, session='0')

Write the paradigm to a csv file

Parameters:

csv_file: string, path of the csv file

session: string, optional, session identifier

Function

nipy.modalities.fmri.experimental_paradigm.load_paradigm_from_csv_file(path, session=None)

Read a (.csv) paradigm file consisting of values yielding (occurrence time, (duration), event ID, modulation) and returns a paradigm instance or a dictionary of paradigm instances

Parameters:

path: string,

path to a .csv file that describes the paradigm

session: string, optional, session identifier

by default the output is a dictionary of session-level dictionaries indexed by session

Returns:

paradigm, paradigm instance (if session is provided), or

dictionary of paradigm instances otherwise, the resulting session-by-session paradigm

Notes

It is assumed that the csv file contains the following columns: (session id, condition id, onset), plus possibly (duration) and/or (amplitude). If all the durations are 0, the paradigm will be handled as event-related.

FIXME: would be much clearer if amplitude was put before duration in the .csv