prophet

Tool for producing high quality forecasts for time series data that has multiple seasonality with linear or non-linear growth.

by facebook

forecasting python r

Star on GitHub ForkWebsite

Python

20.3k stars 4.6k forks 191 contributorsActive · 1mo agoSince 2016 v1.3.0MIT

Meet the team

See all 191 on GitHub →

bletham370 contributions

tcuongd65 contributions

seanjtaylor28 contributions

ryankarlos23 contributions

Bot

dependabot[bot]20 contributions

WardBrian10 contributions

baogorek6 contributions

jorenham6 contributions

Languages

View on GitHub →

Python61.5%

R36.7%

Stan1.8%

Dockerfile0%

Makefile0%

C++0%

Commit activity

Last 12 weeks · 0 commits

Full graph →

Community health

4 of 6 standards met

Community profile →

✓README✓License✓Contributing✓Code of Conduct○Issue Template○PR Template

Recent PRs & issues

Active · Last activity 1mo ago

See all on GitHub →

Warn when yearly seasonality is forced on with insufficient history (#2709)OpenPR

Fixes #2709. When is set explicitly on under ~2 years of data, the yearly component is under-identified and the trend/seasonality split becomes unstable across Prophet/Stan versions (the linked issue shows the same data producing +2.9M vs -12.4M forecasts depending on environment). The instability itself is fundamental to the math, so this doesn't try to "fix" the forecasts. Instead it adds the safeguard the reporter asked for: a in when yearly seasonality is forced on with insufficient history, so users get a clear heads-up rather than unstable output. The warning fires only when seasonality is forced on with under 730 days of data. It stays silent on (which already disables yearly seasonality with its own info-level message) and when there are >= 2 years of data. Includes two tests in covering both the warning case and the no-warning-on-auto case.

anp-exe · 3w ago

Fix unconditional fig.tight_layout() warnings (Issue #2257)OpenPR

Fixes facebook/prophet#2257 by conditionally checking the layout engine before calling to prevent warnings when is used.

amanv0007 · 1mo ago

Update outdated license info in python init file (Fixes #2728)OpenPR

Fixes facebook/prophet#2728 by updating the old BSD-style license header in to match the MIT license used in the rest of the project.

amanv0007 · 1mo ago

Recent fixes

View closed PRs →

Fix: Warn when yearly seasonality is forced on with insufficient history (#2709)MergedPR

This PR fixes #2736. Synthesized autonomously by GitGrok, an AI agent built by @harshagm665-netizen.

harshagm665-netizen · 1w ago

fix(forecaster): correct 'occured' -> 'occurred' in slope-delta helper docstringMergedPR

Docstring inside the slope-delta matrix helper at used . Doc-only change; stays clean against current .

SAY-5 · 2mo ago

NumPy dependency/requirement should be ≥1.20ClosedIssue

Currently, the prophet Python package dependency requirement is . However, older versions of NumPy cause issues when tries to do . The NumPy version issue regarding the submodule is documented here, here, and here, among others. If is going to import from (as it does in ), it should require Numpy v1.20 or greater.

cgobat · 2mo ago

Structured data for AI agents

Repository: facebook/prophet. Description: Tool for producing high quality forecasts for time series data that has multiple seasonality with linear or non-linear growth. Stars: 20271, Forks: 4632. Primary language: Python. Languages: Python (61.5%), R (36.7%), Stan (1.8%), Dockerfile (0%), Makefile (0%). License: MIT. Homepage: https://facebook.github.io/prophet Topics: forecasting, python, r. Latest release: v1.3.0 (5mo ago). Open PRs: 12, open issues: 452. Last activity: 1mo ago. Community health: 75%. Top contributors: bletham, tcuongd, seanjtaylor, ryankarlos, dependabot[bot], WardBrian, baogorek, jorenham, seriousran, joseangel-sc and others.