Skip to contents

Generate power spectral density (PSD) estimate using Welch’s method

psd.Rd

This function mimics the PSD estimation behavior of PyCBC's `welch()` function within an R pipeline. It computes the power spectral density of a time series by applying the Welch method—segmenting the data, applying a window, taking the median average (robust to outliers), then interpolating and truncating the inverse spectrum as needed.

Usage

psd(ts, sl = 4, fl = 15, delf = NULL, window_func = bspec::hannwindow)

Arguments

ts

A `ts` time-series object.

sl

Numeric scalar. Segment length for the Welch method (in time units).

fl

Numeric scalar. Low-frequency cutoff for filter truncation (same units as `ts` frequency).

delf

Numeric scalar, optional. Frequency resolution; defaults to `frequency(ts) / length(ts)`.

window_func

Function. A window function (e.g., `bspec::hannwindow`) to shape each segment prior to PSD estimation.

Value

An object of class `fs`, representing the estimated PSD: - rows by frequency bin, - interpolated to align with spacing `delf`, and - processed via inverse-spectrum truncation at `fl`.

Details

The core estimation pipeline proceeds as follows: 1. **Welch PSD estimation** using `bspec::welchPSD()`, with median averaging for robustness. 2. Conversion to `fs` object via `fs()`, pairing PSD power with `frequency(ts)`-based spacing. 3. **Interpolation** to uniform frequency resolution via `interp_psd()`. 4. **Inverse-spectrum truncation** using `inv_spec_trunc_psd()` to filter low-frequency components below `fl` using the provided window.

This mirrors the standard usage of `pycbc.psd.welch()` for PSD estimates via Welch’s method in gravitational-wave data analysis. [oai_citation:1‡PyCBC](https://pycbc.org/pycbc/latest/html/_modules/pycbc/psd/estimate.html)

References

- PyCBC `welch()` documentation via Sphinx: input parameters `seg_len`, `window`, `avg_method='median'`, returns `FrequencySeries` PSD estimates. [oai_citation:2‡PyCBC](https://pycbc.org/pycbc/latest/html/_modules/pycbc/psd/estimate.html) - General methodology: Welch (1967), The use of smoothing windows in spectral estimation.