RATSpy

RATSpy provides Python bindings for RATS, a rust crate for Rapid Augmentations for Time Series on univariate time series data.

Installation

Install the package using PyPI

pip install ratspy

Python version 3.8+ is required.

Usage

Example to reconstruct the surface from an input file, apply some post-processing methods and write the data back to a file:

import ratspy as rp
import numpy as np

# Dataset from https://timeseriesclassification.com/dataset.php
data = pd.read_csv("./data/Car/Car.csv").to_numpy()

x = data[:,:-1].astype(np.float64)
y = list(map(lambda a: str(a), data[:,-1]))

dataset = rp.Dataset(x, y)

addnoise = rp.AddNoise(rp.NoiseType.Slope, bounds=(0.01, 0.05))
# Only execute the AddNoise augmenter for half of the series in the dataset
addnoise.probability = 0.5

pipeline = (rp.AugmentationPipeline()
            + rp.Repeat(10)
            + rp.Crop(100)
            + addnoise
            + rp.Jittering(0.1))

pipeline.augment_batch(dataset, parallel=True)

# Access augmented data using dataset.features and dataset.labels

Development notes

Manual Build instructions

1. Clone the repository
2. cd to the ratspy directory
3. Install all required Python packages from requirements.txt in a virtual environment of your choice: pip install -r requirements.txt
4. Now, to build the project, use maturin: maturin develop -r
   • Maturin automatically installs the resulting binary in your Python environment

Stub file generation

To automatically generate a stub file (ratspy.pyi) for the package, run cargo run --bin stub_gen.

This will expose Rust documentation comments to the Python side as docstrings.

Sphinx Documentation

To generate the Sphinx documentation, make sure that the package is installed through, e.g., maturin, and then run make html in the docs directory. The resulting HTML files will be in docs/_build/html.

You also need to install the documentation dependencies from docs/requirements.txt.

Tests

To test the installation of ratspy in your venv and run some initial tests, we provide unit tests for each of ratspy's classes as well as the entire pipeline in the /tests directory.

Run the following commands to run tests:

cd tests
python -m unittest test_RATSpy

Benchmarking

In benchmarking/ are scripts that automatically benchmark this library against the Python-native library tsaug. Which augmenters should be benchmarked can be configured using the benchmarking/augmenter_configs.yaml file.