acfmperiod

acfmperiod

Python port of the original R package acfMPeriod for classical and robust autocovariance/autocorrelation estimation using periodogram and M-periodogram methods.

Features

• Univariate and multivariate ACF/ACOVF from periodogram (PerACF)
• Robust ACF/ACOVF from M-periodogram (MPerACF)
• Periodogram and cross-periodogram estimators
• Covariance/correlation matrix wrappers at lag 0 (CovCorPer, CovCorMPer)
• Plotting support for robust and classical outputs

Installation

pip install acfmperiod

For local development:

pip install -e .[dev]

Quick Start

import numpy as np
from acfmperiod import PerACF, MPerACF, CovCorPer

rng = np.random.default_rng(7)
x = rng.normal(size=(128, 2))

res = PerACF(x, lagmax=20, typevalue="correlation", doplot=False)
print(res.acf.shape)   # (20, 2, 2)

rob = MPerACF(x, lagmax=20, doplot=False)
cov0 = CovCorPer(x, typevalue="covariance")
print(cov0)

API

• PerioReg, MPerioReg
• CrossPeriodogram, MCrossPeriodogram
• PerACF, MPerACF
• CovCorPer, CovCorMPer
• plotrobacf

Lowercase aliases are available without underscore naming. ß