analysis.granger

Module: analysis.granger

Inheritance diagram for nitime.analysis.granger:

Analyzers for the calculation of Granger ‘causality’

GrangerAnalyzer

class nitime.analysis.granger.GrangerAnalyzer(input=None, ij=None, order=None, max_order=10, criterion=<function bayesian_information_criterion>, n_freqs=1024)

Bases: nitime.analysis.base.BaseAnalyzer

Analyzer for computing all-to-all Granger ‘causality’

__init__(input=None, ij=None, order=None, max_order=10, criterion=<function bayesian_information_criterion>, n_freqs=1024)

Initializer for the GrangerAnalyzer.

Parameters:

input: nitime TimeSeries object :

ij: List of tuples of the form: [(0, 1), (0, 2)], etc. :

These are the indices of pairs of time-series for which the analysis will be done. Defaults to all vs. all.

order: int (optional) :

The order of the process. If this is not known, it will be estimated from the data, using the information criterion

max_order: if the order is estimated, this is the maximal order to :

estimate for.

n_freqs: int (optional) :

The size of the sampling grid in the frequency domain. Defaults to 1024

criterion: :

XXX

autocov()

causality_xy()

causality_yx()

error_cov()

frequencies()

model_coef()

order()

simultaneous_causality()

spectral_matrix()

nitime.analysis.granger.fit_model(x1, x2, order=None, max_order=10, criterion=<function bayesian_information_criterion>)

Fit the auto-regressive model used in calculation of Granger ‘causality’.

Parameters:

x1,x2: float arrays (n) :

x1,x2 bivariate combination.

order: int (optional) :

If known, the order of the autoregressive process

max_order: int (optional) :

If the order is not known, this will be the maximal order to fit.

criterion: callable :

A function which defines an information criterion, used to determine the

order of the model.