src.spenv

Attributes

__version__

Functions

spec_taper(x[, p])

Apply a cosine-bell taper to a time series.

mvspec(x[, xfreq, spans, kernel, taper, pad, fast, ...])

Computes the spectral density estimate using a periodogram.

specenv(xdata[, xfreq, spans, kernel, taper, pad, ...])

Spectral envelope of real-valued time series.

spec_opt(x, *args[, as_series])

Automatically smoothes a time series using the Spectral Envelope method.

Module Contents

src.spenv.__version__
src.spenv.spec_taper(x, p=0.1)

Apply a cosine-bell taper to a time series.

Computes a tapered version of x, with tapering proportion p at each end of x.

Parameters:

  • x (float) – A univariate or multivariate time series

  • p (float) – The proportion to be tapered at each end of the series, either a scalar (giving the proportion for all series) or a vector of the length of the number of series (giving the proportion for each series).

Returns:

A new time series.

Return type:

x (float)

Adapted from R’s stats::spec.taper.

src.spenv.mvspec(x, xfreq=1, spans=None, kernel=None, taper=0.1, pad=0, fast=True, demean=False, detrend=True, minimal=True, option_summary=False, **kwargs)

Computes the spectral density estimate using a periodogram.

Parameters:

  • x (numpy array) – Univariate or multivariate time series.

  • xfreq (optional) – Number of samples per unit time. Defaults to 1.

  • spans (optional) – Sequence of spans for convoluted Daniell smoothers. Defaults to None.

  • kernel (optional) – Defines Kernel for smoothing. Defaults to None.

  • taper (optional) – Defines proportion for tapering start and end of series to avoud end-of-signal effects. Defaults to 0.1.

  • pad (optional) – Pads the provided series before computation, adding pad*(length of series) zeros at the end. Defaults to 0.

  • fast (optional) – [description]. Defaults to True.

  • demean (optional) – Demeans series. Defaults to False.

  • detrend (optional) – Detrends series. Defaults to True.

  • minimal (optional) – Returns only frequency and spectrum. Overrides option_summary. Defaults to True.

  • option_summary (optional) – Returns specified options alongside results. Defaults to False.

Adapted from R’s astsa::mvspec and the spec_pgram function in https://github.com/kircherlab/cfDNA.

src.spenv.specenv(xdata, xfreq=1, spans=None, kernel=None, taper=0, pad=0, fast=True, demean=False, detrend=True, **kwargs)

Spectral envelope of real-valued time series.

Computes the Spectral Envelope as described in Time Series Analysis and Its Applications with R Examples and the R package ASTSA.

src.spenv.spec_opt(x, *args, as_series=False)

Automatically smoothes a time series using the Spectral Envelope method.

Parameters:

  • x (numpy array) – Univariate time series.

  • *args (optional) – Mapping functions for the Spectral Envelope.