OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

River

River is an open-source Python library for online machine learning: it processes observations individually so developers can make predictions and update models as streaming data arrives. It includes algorithms and utilities for classification, anomaly and drift detection, forecasting, clustering, preprocessing, metrics, pipelines and progressive validation.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

River — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use River when a model should learn from new observations as they arrive, or when an event-based workflow and changing data patterns make incremental learning useful. Batch learning may be a better fit when continual updates are unnecessary.

Who it is for

  • Python developers building machine-learning workflows around continuously arriving records.
  • Teams that want models to learn incrementally without revisiting all prior data.
  • Practitioners addressing streams where the relationship between data and outcomes may shift over time.

Documented uses

  • Predict outcomes for incoming individual records and update a model from new observations.
  • Build online classification, anomaly-detection, forecasting or clustering workflows.
  • Track online model performance with metrics and progressive validation.
  • Add online preprocessing and pipeline components to a streaming-learning workflow.
Before you choose

Requirements & limits

Open-source Python library under the BSD-3-Clause license. The project specifies Python 3.11 or later and documents pip installation; wheels are listed for Linux, macOS and Windows.

  • The core online interface does not depend on pandas; mini-batch methods use pandas and are installed as an optional extra.
  • The project emphasizes clarity and usability over performance, while describing single-sample processing as fast.
  • Online learning is useful when data arrives incrementally or concept drift matters; the project notes that batch learning is sufficient for many cases.
  • A GitHub development-version install requires Cython and Rust.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install River for single-observation learning

Set up the library for Python-based online prediction and model updates.

Before you start

  • Python 3.11 or later
  • pip

Steps

  1. Install the package with the command below.

    Documented command
    pip install river

  2. In a Python workflow, use River’s predict_one and learn_one interfaces to predict from and update with individual observations.

Expected outcome

River is installed for a single-observation online-learning workflow; the core interfaces do not require pandas.

Guide limitations

  • The installation step alone does not configure a dataset, model, or production data pipeline.
  • Mini-batch interfaces require the opt-in pandas extra.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
10 October 2026
Original website title
online-ml/river · Official homepage title · retained in its source language
How we review sourcesSuggest a correction