Python-Implementation-of-ARMORF-Function-from-MATLAB-Toolbox-BSMART

Python Implementation of ARMORF Function from MATLAB Toolbox BSMART

ARMORF can be used to compute autoregressive factors and to measure Granger causality.

Reference:

M. Morf, etal, Recursive Multichannel Maximum Entropy Spectral Estimation, IEEE trans. GeoSci. Elec., 1978, Vol.GE-16, No.2, pp85-94.

S. Haykin, Nonlinear Methods of Spectral Analysis, 2nd Ed. Springer-Verlag, 1983, Chapter 2

Jie Cui, Lei Xu, Steven L. Bressler, Mingzhou Ding, Hualou Liang, BSMART: a Matlab/C toolbox for analysis of multichannel neural time series, Neural Networks, 21:1094 - 1104, 2008.

:unicorn:How to use

```Python import numpy as np from matplotlib import pyplot as plt from armorf imoirt armorf

Create a toy series

x = [1]
for i in range(30): x.append(1.1 * x[-1])

Compute the AR factor

Ax, Ex = armorf(np.asarray(x).reshape(1, -1),1,len(x),2)

Fit a new series using AR factor

y = [1, 1.1] factor_0, factor_1 = np.squeeze(Ax) for i in range(30): y.append(factor_0 * y[-1] + factor_1 * y[-2])

plt.figure() plt.clf() plt.plot(np.asarray(x)-0.5, label=’x’) # Adjust the intercept of x for better visualization plt.plot(y, label=’y’) plt.legend() plt.show() ```

This site is open source. Improve this page.