OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Data & Analytics

Statsforecast

Statsforecast is an open-source Python library that fits statistical and econometric models to univariate time series and generates forecasts.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Statsforecast — OpenViST project profile in AI Data & Analytics.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when a Python workflow needs statistical time-series forecasts, model comparisons or the documented forecasting-related tasks such as intermittent-demand analysis and electricity peak forecasting.

Who it is for

  • Python developers building forecasting into analytics or production workflows.
  • Analysts and data scientists comparing statistical models across time series.
  • Teams working with large collections of series or using Spark, Dask or Ray.

Documented uses

  • Generate forecasts with automatic ARIMA, ETS, CES, Theta and other benchmark models.
  • Evaluate and select models across multiple time series.
  • Request probabilistic forecasts and confidence intervals, including ARIMA forecasts with exogenous variables.
  • Detect anomalies using in-sample prediction intervals.
Before you choose

Requirements & limits

The repository lists an Apache-2.0 license and installation through pip or conda-forge.

  • It is a Python library with fit/predict-style usage, not a separately documented hosted forecasting service.
  • The repository lists Spark, Dask and Ray compatibility for distributed workflows.
  • Its model collection includes automatic ARIMA, ETS, CES and Theta approaches as well as benchmarking models.

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 StatsForecast with pip

Install the released StatsForecast package in a Python environment.

Before you start

  • Python package index access; a Python virtual environment or conda environment is recommended.

Steps

  1. Optionally activate a Python virtual environment or conda environment.

  2. Run the command below to install the released package.

    Documented command
    pip install statsforecast

Expected outcome

StatsForecast is installed from the Python package index.

Guide limitations

  • The installation guide recommends fixing the package version to avoid issues but does not establish which version to use for a particular project.
  • Installing the development version from GitHub requires a C++ compiler; this recipe uses the released package instead.
Documentation 1 ↗
Source-based practical guide

Generate an AirPassengers forecast with AutoARIMA

Fit an AutoARIMA model to the included AirPassengers data and request a 12-step forecast with a 95% prediction level.

Before you start

  • StatsForecast installed in a Python environment.

Steps

  1. Import StatsForecast, AutoARIMA, and AirPassengersDF from their StatsForecast modules.

  2. Set df = AirPassengersDF and create StatsForecast(models=[AutoARIMA(season_length=12)], freq='ME').

  3. Call sf.fit(df).

  4. Call sf.predict(h=12, level=[95]) to request the forecast.

Expected outcome

The library returns the requested forecast from the fitted AutoARIMA model.

Guide limitations

  • The example uses the included AirPassengers dataset and does not document adapting the code to other datasets.
  • The README does not specify a calendar interpretation for the ME frequency or further details about prediction-level semantics.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
10 October 2026
Original website title
Nixtla/statsforecast · Official homepage title · retained in its source language
How we review sourcesSuggest a correction