Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/predict-idlab/tsflex
Flexible time series feature extraction & processing
https://github.com/predict-idlab/tsflex
data-science feature-engineering feature-extraction multimodal multivariate pandas processing python time-series window-stride
Last synced: 2 months ago
JSON representation
Flexible time series feature extraction & processing
- Host: GitHub
- URL: https://github.com/predict-idlab/tsflex
- Owner: predict-idlab
- License: mit
- Created: 2021-07-16T14:23:32.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2024-04-19T06:57:51.000Z (9 months ago)
- Last Synced: 2024-04-19T08:55:18.547Z (9 months ago)
- Topics: data-science, feature-engineering, feature-extraction, multimodal, multivariate, pandas, processing, python, time-series, window-stride
- Language: Python
- Homepage: https://predict-idlab.github.io/tsflex/
- Size: 28 MB
- Stars: 360
- Watchers: 7
- Forks: 25
- Open Issues: 31
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- Funding: .github/FUNDING.yml
- License: LICENSE
Awesome Lists containing this project
- awesome-time-series - tsflex: Flexible Time Series Processing & Feature Extraction
- awesome_time_series_in_python - tsflex
- awesome-time-series - tsflex
- awesome-TS-anomaly-detection - tsflex - window feature extraction on multivariate, irregularly-sampled sequence data. | MIT | :heavy_check_mark: (Related Software / Time-Series Analysis)
README
#
[](https://pypi.org/project/tsflex/)
[](https://anaconda.org/conda-forge/tsflex)
[](https://img.shields.io/pypi/pyversions/tsflex)
[](https://codecov.io/gh/predict-idlab/tsflex)
[](https://github.com/predict-idlab/tsflex/actions/workflows/codeql.yml)
[](https://pepy.tech/project/tsflex)
[](http://makeapullrequest.com)
[](https://github.com/predict-idlab/tsflex/actions/workflows/deploy-docs.yml)
[](https://github.com/predict-idlab/tsflex/actions/workflows/test.yml)
[](https://discord.gg/k2d59GrxPX)
> _tsflex_ is a toolkit for _**flex**ible **t**ime **s**eries_ [processing](https://predict-idlab.github.io/tsflex/processing) & [feature extraction](https://predict-idlab.github.io/tsflex/features), that is efficient and makes few assumptions about sequence data.
#### Useful links
- [Paper](https://www.sciencedirect.com/science/article/pii/S2352711021001904)
- [Documentation](https://predict-idlab.github.io/tsflex/)
- [Example (machine learning) notebooks](https://github.com/predict-idlab/tsflex/tree/main/examples)
#### Installation
| | command |
| :--------------------------------------------------- | :------------------------------------ |
| [**pip**](https://pypi.org/project/tsflex/) | `pip install tsflex` |
| [**conda**](https://anaconda.org/conda-forge/tsflex) | `conda install -c conda-forge tsflex` |
## Usage
_tsflex_ is built to be intuitive, so we encourage you to copy-paste this code and toy with some parameters!
```python
import pandas as pd; import numpy as np; import scipy.stats as ss
from tsflex.features import MultipleFeatureDescriptors, FeatureCollection
from tsflex.utils.data import load_empatica_data
# 1. Load sequence-indexed data (in this case a time-index)
df_tmp, df_acc, df_ibi = load_empatica_data(['tmp', 'acc', 'ibi'])
# 2. Construct your feature extraction configuration
fc = FeatureCollection(
MultipleFeatureDescriptors(
functions=[np.min, np.mean, np.std, ss.skew, ss.kurtosis],
series_names=["TMP", "ACC_x", "ACC_y", "IBI"],
windows=["15min", "30min"],
strides="15min",
)
)
# 3. Extract features
fc.calculate(data=[df_tmp, df_acc, df_ibi], approve_sparsity=True)
```
Note that the feature extraction is performed on multivariate data with varying sample rates.
| signal | columns | sample rate |
|:-------|:-------|------------------:|
| df_tmp | ["TMP"]| 4Hz |
| df_acc | ["ACC_x", "ACC_y", "ACC_z" ]| 32Hz |
| df_ibi | ["IBI"]| irregularly sampled |
### Processing
[Working example in our docs](https://predict-idlab.github.io/tsflex/processing/index.html#working-example)
## Why tsflex? ✨
- `Flexible`:
- handles multivariate/multimodal time series
- versatile function support
=> **integrates** with many packages for:
- processing (e.g., [scipy.signal](https://docs.scipy.org/doc/scipy/reference/tutorial/signal.html), [statsmodels.tsa](https://www.statsmodels.org/stable/tsa.html#time-series-filters))
- feature extraction (e.g., [numpy](https://numpy.org/doc/stable/reference/routines.html), [scipy.stats](https://docs.scipy.org/doc/scipy/reference/tutorial/stats.html), [antropy](https://raphaelvallat.com/antropy/build/html/api.html), [nolds](https://cschoel.github.io/nolds/nolds.html#algorithms), [seglearn](https://dmbee.github.io/seglearn/feature_functions.html)¹, [tsfresh](https://tsfresh.readthedocs.io/en/latest/text/list_of_features.html)¹, [tsfel](https://tsfel.readthedocs.io/en/latest/descriptions/feature_list.html)¹)
- feature extraction handles **multiple strides & window sizes**
- `Efficient`:
- view-based operations for processing & feature extraction => extremely **low memory peak** & **fast execution time**
- see: [feature extraction benchmark visualization](https://predict-idlab.github.io/tsflex/#benchmark)
- `Intuitive`:
- maintains the sequence-index of the data
- feature extraction constructs interpretable output column names
- intuitive API
- `Few assumptions` about the sequence data:
- no assumptions about sampling rate
- able to deal with multivariate asynchronous data
i.e. data with small time-offsets between the modalities
- `Advanced functionalities`:
- apply [FeatureCollection.**reduce**](https://predict-idlab.github.io/tsflex/features/index.html#tsflex.features.FeatureCollection.reduce) after feature selection for faster inference
- use **function execution time logging** to discover processing and feature extraction bottlenecks
- embedded [SeriesPipeline](http://predict-idlab.github.io/tsflex/processing/#tsflex.processing.SeriesPipeline.serialize) & [FeatureCollection](https://predict-idlab.github.io/tsflex/features/index.html#tsflex.features.FeatureCollection.serialize) **serialization**
- time series [**chunking**](https://predict-idlab.github.io/tsflex/chunking/index.html)
¹ These integrations are shown in [integration-example notebooks](https://github.com/predict-idlab/tsflex/tree/main/examples).
## Future work 🔨
- scikit-learn integration for both processing and feature extraction
**note**: is actively developed upon [sklearn integration](https://github.com/predict-idlab/tsflex/tree/sklearn_integration) branch.
- Support time series segmentation (exposing under the hood strided-rolling functionality) - [see this issue](https://github.com/predict-idlab/tsflex/issues/15)
- Support for multi-indexed dataframes
=> Also see the [enhancement issues](https://github.com/predict-idlab/tsflex/issues?q=is%3Aissue+is%3Aopen+label%3Aenhancement+)
## Contributing 👪
We are thrilled to see your contributions to further enhance `tsflex`.
See [this guide](CONTRIBUTING.md) for more instructions on how to contribute.
## Referencing our package
If you use `tsflex` in a scientific publication, we would highly appreciate citing us as:
```bibtex
@article{vanderdonckt2021tsflex,
author = {Van Der Donckt, Jonas and Van Der Donckt, Jeroen and Deprost, Emiel and Van Hoecke, Sofie},
title = {tsflex: flexible time series processing \& feature extraction},
journal = {SoftwareX},
year = {2021},
url = {https://github.com/predict-idlab/tsflex},
publisher={Elsevier}
}
```
Link to the paper: https://www.sciencedirect.com/science/article/pii/S2352711021001904
---
👤 Jonas Van Der Donckt, Jeroen Van Der Donckt, Emiel Deprost