# WenjieDu/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

Repository: https://github.com/WenjieDu/PyPOTS
Canonical: https://ross.abutalabs.com/products/pypots
Homepage: https://pypots.com
Language: Python
License: BSD-3-Clause
License Family: permissive
Topics: time-series, data-mining, deep-learning, missing-values, anomaly-detection, classification, clustering, forecasting, generation, imputation, machine-learning, pytorch, data-analysis, data-science, neural-networks
Last push: 2026-08-24T17:54:16+00:00

## Health v2 (maintenance only)
Score: 93/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 82, longevity 100
- inputs: {"age_days": 1618, "days_push": 9, "days_rel": 120, "gap_med": 15.5, "n_releases_24m": 19}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2051, forks 188 (observed 2026-08-28T04:06:09.206949+00:00)

## What it is
PyPOTS is a Python toolbox for machine learning and data mining on partially-observed time series with missing values. It integrates 50+ state-of-the-art neural network models for imputation, classification, clustering, forecasting, and anomaly detection with a unified fit/predict API built on PyTorch.

## Use cases
- impute missing values in multivariate time series with NaNs
- forecast irregularly-sampled time series with missing data
- classify partially-observed time series
- cluster incomplete time series data
- detect anomalies in sensor data with missing readings
- benchmark SOTA models on incomplete time series datasets

## When to choose
- your time series data has missing values (NaNs) from sensor failures or irregular sampling
- you want a unified scikit-learn-like API to try many SOTA imputation/forecasting/classification models
- you work in healthcare, IoT, climate, or industrial domains with incomplete multivariate time series
- you want reproducible, peer-reviewed implementations with benchmarks

## When to avoid
- your time series data is fully observed with no missing values
- you need streaming/real-time inference rather than batch analysis
- you need a framework outside Python/PyTorch
- you need classical statistical methods only (e.g., simple ARIMA) without deep learning

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, data-science, nlp
- domain: machine-learning, data-science, time-series, artificial-intelligence
- platform: python
- tags: time-series, missing-data, imputation, anomaly-detection, forecasting, pytorch, partially-observed-time-series, clustering, classification

## Member repositories
- WenjieDu/PyPOTS (main) score 93

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:09.206949+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:57:30.649100+00:00, confidence not recorded.
  - readme: https://github.com/WenjieDu/PyPOTS (fetched 2026-08-28T04:06:09.206949+00:00, sha c5039a1b90f0)
  - homepage: https://pypots.com (fetched 2026-08-29T10:38:26.319697+00:00, sha d1a5ae3f4ff8)
  - site_page: https://pypots.com/about (fetched 2026-08-29T10:38:26.329062+00:00, sha b3a4ff52e93b)
  - registry_pypi: https://pypi.org/pypi/pypots/json (fetched 2026-08-29T10:38:26.332751+00:00, sha f6c2209a2fd7)
- Data as of 2026-08-30T08:39:29.467469+00:00.