Spectral Analysis Library (Welch's Method)

📖 Overview

This library provides an implementation of Welch's method for estimating the Power Spectral Density (PSD) and Power Spectrum of a time-series signal. Welch's method is a non-parametric technique that reduces the variance of the standard periodogram by averaging the modified periodograms of overlapping signal segments.

The core functionality is exposed through public wrappers for f32 and f64 data types.

✨ Key Features

• Data Types: Supports both single-precision (f32) and double-precision (f64) input.
• Windowing: Multiple window functions are available to minimize spectral leakage:
  • WelchWindow::Hann
  • WelchWindow::Hamming
  • WelchWindow::Blackman
• Detrending: Supports removing DC offsets (DetrendingMethod::Constant) or linear trends (DetrendingMethod::Linear) from each segment.
• Scaling Control: Configure the output as either:
  • Power Spectral Density (Density): Units of Power/Hz.
  • Power Spectrum (Spectrum): Units of Power.

⚙️ Public Functions and Parameters

The primary entry points are the type-specific functions: welch_f32 and welch_f64.

This project is licensed under either of

• BSD-3-Clause License (see LICENSE)
• Apache License, Version 2.0 (see LICENSE)

at your option.