pyDySP 0.2.0

Python Dynamic Signal Processing

pyDySP

Dynamic Signal Processing for Experimental Data

pyDySP is a lightweight Python library for dynamic signal processing of experimental data, developed for
the EQUALS and SoFSI Laboratories at the University of Bristol.

It provides:

• A Channel class for time-history signals
  (processing, spectra, trimming, Arias, response spectra, plots, metadata).
• A Test class for multi-channel experiments
  (selection, batch processing, cross-spectra, transfer functions, modal analysis via sdypy, plotting utilities, .mat/.csv I/O).
• Clean, readable, notebook-friendly code for research, teaching and reproducible workflows.
• Full internal documentation and example notebooks.

Table of Contents

• pyDySP
  • Table of Contents
  • Installation
    • Option A — Install from PyPI (recommended)
    • Option B — Developer installation (editable)
    • Option C — Use directly from a folder
  • Environment Setup
    • Virtual environment (venv)
    • Conda / Mamba environment
  • Dependencies
  • Quick Start
    • Single channel
    • Multi-channel test
    • Spectra & trimming
    • Response spectrum
    • Transfer functions
    • Channel health
    • Modal identification (optional)
  • Features
    • Channel
    • Test
  • Example Notebooks
  • Documentation
  • Contributing
  • License

Installation

Option A — Install from PyPI (recommended)

pip install pydysp

For Jupyter notebooks:

pip install ipykernel

Option B — Developer installation (editable)

git clone https://github.com/dkaramitros/pyDySP.git
cd pyDySP
pip install -e .

Option C — Use directly from a folder

import sys
sys.path.append("/path/to/parent/of/pydysp")
from pydysp import Channel, Test

Environment Setup

Virtual environment (venv)

python -m venv .venv
source .venv/bin/activate      # Linux/Mac
.venv\Scripts\activate         # Windows
pip install pydysp

Conda / Mamba environment

mamba create -n pydysp python=3.11 pip
mamba activate pydysp
pip install pydysp

Dependencies

Required:

• numpy
• scipy
• matplotlib
• tabulate

Optional (for modal analysis):

pip install sdypy-EMA

Quick Start

Single channel

from pydysp import Channel
import numpy as np

dt = 0.01
t = np.arange(0, 10, dt)
y = np.sin(2*np.pi*2*t)

ch = Channel(data=y, dt=dt, name_user="Acc1", quantity="acceleration", units="g")
ch.plot()
t_proc, y_proc = ch.processed()

Multi-channel test

from pydysp import Test

test = Test.from_channels(name="MyTest", channels=[ch1, ch2])
print(test.info())
test.plot_channels(ncols=2)

Spectra & trimming

spec = ch.fourier()
f_peak, s_peak = spec.peak()

ch_trim = ch.trim_by_arias()

Response spectrum

rs = ch.response_spectrum()
rs.plot(y="Sa", logx=True)

Transfer functions

f, H = test.transfer_function("Shaker", "Acc1", kind="H1")

Channel health

print(test.channel_health())

Modal identification (optional)

model = test.ema_model(input="Shaker", outputs=["Acc1","Acc2"], lower=2, upper=40)
model.get_poles()
model.select_poles()
model.print_modal_data()

Features

Channel

• Drift correction, filtering, baseline correction
• Trimming (threshold, fraction, Arias)
• Fourier & Welch spectra
• Arias intensity
• Response spectrum
• Time-domain metrics
• Metadata support
• Plot utilities

Test

• Channel selection (by index, name, tags, slices)
• Batch processing
• Cross-spectra, transfer functions, delays
• Modal analysis (via sdypy)
• .mat (SoFSI/EQUALS) and .csv I/O
• Grid plots & channel summary tables
• Health diagnostics

Example Notebooks

Located in:

examples/

Documentation

https://pydysp.readthedocs.io/en/latest/

Contributing

Contributions are very welcome — whether bug reports, small fixes, feature ideas, or full pull requests.

To contribute:

1. Fork the repository on GitHub

2. Create a branch for your change

3. Commit your edits with clear messages

4. Run the tests locally:

   pip install -e .
   pip install pytest
   pytest -q
   

5. Open a Pull Request describing the change and its motivation

Please keep code style readable and consistent with the existing project
(docstrings, small focused functions, and meaningful error messages).

If you prefer to discuss an idea before writing code, feel free to open a GitHub Issue.

License

MIT License