Welcome to PulsarSA documentation!

Welcome to PulsarSA! This guide provides all the information you need to install and run the pulsarsa Python package. For a quick start, we recommend beginning with the Getting Started page, followed by the Examples page. For more in-depth usage, please refer to the example notebooks in the examples folder. Documentation related to the source code can be found on the Modules page.

Below is a simple diagram of PulsarSA functionalities:

PipelineImageToFilterToCCtoLabels

Here are the list of modules used to perform different steps in the pipeline:

  • Loaders: Data loaders and converters (a) from PulsarDT payload files

  • Datasets: Data sets for training, testing, and visualization (b), (c)

  • Trainers: Trainer modules (d), (e) to train CNNs (f) and (g) for segmentation and filtering

  • Labellers: Legacy vision tools (h) to label segments into different signal source types (Pulsar, NBRFI, BBRFI)

  • Optimizers: Optimizers (j) to find the best model parameters mix using the Tuner (i) to match the real world data

Contents

Getting Started

Learn how to install and use the library.

getting_started.html
Examples

See real-world examples and demos.

examples.html
Modules

Explore the main modules and functions.

modules.html
Tools / Utilities

Functions for easy to use pipeline functionalities.

src.pulsarsa.tools.html

Acknowledgements

We gratefully acknowledge the contributions of the following individuals and institutions:

  • HTW Berlin and PUNCH4NFDI for project support

  • Prof. Dr. Hermann Heßling for guidance

  • Andrei Kazantsev for valuable feedbacks and for providing real data

Collaborations

This project is a joint effort between:

  • HTW Berlin, Department of Computer Science

  • PUNCH4NFDI

  • MPIfRBonn

  • interTwin

  • DZA

The code is hosted on GitLab and can be found at:

🔗 View on GitLab