# winedarksea/AutoTS

Automated Time Series Forecasting

Repository: https://github.com/winedarksea/AutoTS
Canonical: https://ross.abutalabs.com/products/autots
Language: Python
License: MIT
License Family: permissive
Topics: time-series, machine-learning, automl, autots, forecasting, deep-learning, preprocessing, feature-engineering
Last push: 2026-08-25T22:42:53+00:00

## Health v2 (maintenance only)
Score: 95/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 86, longevity 100
- inputs: {"age_days": 2472, "days_push": 8, "days_rel": 13, "gap_med": 38.5, "n_releases_24m": 11}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1430, forks 122 (observed 2026-08-28T04:04:42.346529+00:00)

## What it is
AutoTS is a Python library for automated time series forecasting, offering dozens of sklearn-style models (statistical, ML, deep learning) plus 30+ time series transforms that operate directly on Pandas DataFrames. It includes an AutoML feature search using genetic algorithms to select models, preprocessing, and ensembles, with support for multivariate and probabilistic forecasts at scale.

## Use cases
- forecast sales or demand across thousands of time series
- automatically find the best forecasting model for my dataset
- generate probabilistic upper and lower bound forecasts
- run an automl search over time series models and preprocessing
- forecast stock market or financial time series
- detect anomalies in time series data
- scale forecasting to hundreds of thousands of input series

## When to choose
- you need automated, high-accuracy time series forecasting in Python with minimal manual tuning
- you work with Pandas DataFrames and want sklearn-style fit/predict APIs
- you need multivariate, probabilistic, or large-scale forecasting with ensembling

## When to avoid
- you need general-purpose AutoML for non-time-series tabular data
- you need a forecasting service with a UI rather than a Python library
- you require real-time streaming forecasting rather than batch prediction

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, data-science, etl
- domain: machine-learning, data-science, time-series, analytics
- platform: python
- tags: time-series, forecasting, automl, sklearn-style, genetic-algorithms, probabilistic-forecasting, anomaly-detection

## Member repositories
- winedarksea/AutoTS (main) score 95

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:42.346529+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 2026-08-30T04:37:13.556507+00:00, confidence not recorded.
  - readme: https://github.com/winedarksea/AutoTS (fetched 2026-08-28T04:04:42.346529+00:00, sha c06fa0aeae4a)
  - registry_pypi: https://pypi.org/pypi/autots/json (fetched 2026-08-29T11:48:46.234243+00:00, sha b43860b0a930)
- Data as of 2026-08-30T08:39:29.467469+00:00.