PyPOTS
A Python toolkit/library for reality-centric machine/deep learning & data mining on partially-observed time series, with 50+ SOTA neural network models for scie
What is PyPOTS?
A Python toolkit/library for reality-centric machine/deep learning & data mining on partially-observed time series, with 50+ SOTA neural network models for scientific analysis tasks (imputation, classification, clustering, forecasting, anomaly detection, cleaning) on incomplete industrial irregularly-sampled multivariate TS with NaN missing values
Overview of PyPOTS
What it is
PyPOTS is a Python toolkit for machine learning on partially-observed time series (POTS). It provides over 50 state-of-the-art neural network models for scientific analysis tasks such as imputation, classification, clustering, forecasting, anomaly detection, and cleaning on incomplete, irregularly-sampled multivariate time series with missing values (NaN). The project aims to fill the lack of a dedicated toolkit for machine learning on POTS data.
Key features
- Supports five core tasks: imputation, forecasting, classification, clustering, and anomaly detection.
- Includes 50+ neural network models, including SAITS, BRITS, Transformer, TimesNet, and many others.
- Provides hyperparameter optimization support via the Optuna framework (since v2.0).
- Offers unified APIs and detailed documentation with interactive examples.
- Includes naive baseline methods (e.g., mean, median, LOCF) and probabilistic models.
- Covers both large language models (LLMs) and time-series foundation models (TSFMs).
Use cases
- Scientific research on partially-observed time series data.
- Industrial applications with missing sensor readings or irregular sampling.
- Benchmarking and comparing imputation, forecasting, or classification algorithms.
- Anomaly detection in multivariate time series with missing values.
Getting started
See the repository README for installation and usage instructions.
License and pricing
PyPOTS is open source under the BSD-3-Clause license.
Based on the project's public README and repository metadata.
