get_lpc

tenpy.tools.prediction.get_lpc(x, p)[source]

Function to obtain the linear prediction coefficients (lpc) from the (last p) correlations of a time series x.

First, the last p correlations are obtained, then the system of equations R x = r (which is in toeplitz form) is solved.

Parameters:

  • x (ndarray) – one-dimensional time series data

  • p (int) – Number of coefficients

Returns:

1D array containing the p linear prediction coefficients [a_p, a_{p-1}, …, a_1] from the last p correlations of the time series x

Return type:

ndarray

Notes

The last p+1 correlations of the one-dimensional time-series are given by their autocorrelation ..math

[E\{x(n)*x(n-0)\}, E\{x(n)*x(n-1)\}, ..., E\{x(n)*x(n-p)\}]