sigpro 0.0.3.dev0

Signal Processing Tools for Machine Mearning

An open source project from Data to AI Lab at MIT.

SigPro: Signal Processing Tools for Machine Learning

• License: MIT
• Development Status: Pre-Alpha
• Homepage: https://github.com/signals-dev/SigPro

Overview

SigPro offers an end-to-end solution to efficiently apply multiple signal processing techniques to convert raw time series into feature time series that encode the knowledge of domain experts in order to solve time series machine learning problems.

Install

Requirements

SigPro has been developed and tested on Python 3.6, 3.7 and 3.8 on GNU/Linux and macOS systems.

Also, although it is not strictly required, the usage of a virtualenv is highly recommended in order to avoid interfering with other software installed in the system where SigPro is run.

Install with pip

The easiest and recommended way to install SigPro is using pip:

pip install sigpro

This will pull and install the latest stable release from PyPi.

If you want to install from source or contribute to the project please read the Contributing Guide.

User Guides

SigPro comes with the following user guides:

• PRIMITIVES.md: Information about the primitive families, their expected input and output.
• USAGE.md: Instructions about how to usee the three main functionalities of SigPro.
• DEVELOPMENT.md: Step by step guide about how to write a valid SigPro primitive and contribute it to either SigPro or your own library.

History

0.0.2 - 2021-02-05

Bug Fixes

• MANIFEST.in: copy the json files of the primitives with the package installation.

0.0.1 - 2021-01-26

First release to PyPI.

This release comes with the first version of the contributing module, which makes it easier to create new primitives and to test those with the demo data included in this package.

This release also includes the following User Guides:

• PRIMITIVES.md: Information about the primitive families, their expected input and output.
• USAGE.md: Instructions about how to usee the three main functionalities of SigPro.
• DEVELOPMENT.md: Step by step guide about how to write a valid SigPro primitive and contribute it to either SigPro or your own library.

Features

• Demo data: Available demo data to test primitives.
• First primitives: The following list of primitives were added:
  • sigpro.aggregations.amplitude.statistical.crest_factor
  • sigpro.aggregations.amplitude.statistical.kurtosis
  • sigpro.aggregations.amplitude.statistical.mean
  • sigpro.aggregations.amplitude.statistical.rms
  • sigpro.aggregations.amplitude.statistical.skew
  • sigpro.aggregations.amplitude.statistical.std
  • sigpro.aggregations.amplitude.statistical.var
  • sigpro.transformations.amplitude.identity.identity
  • sigpro.transformations.frequency.fft.fft
  • sigpro.transformations.frequency.fft.fft_real
  • sigpro.transformations.frequency_time.stft.stft
  • sigpro.transformations.frequency_time.stft.stft_real
• Contributing module.
• Documentation on how to contribute new primitives and how to run those.