skills.rest
JosephWoodall
JosephWoodallCOMMUNITY
Explore all skills by this author →
𝕏@josephwoodall_GitHub

timesfm-forecasting

Forecast univariate time series with quantile prediction intervals.

1|Updated Mar 19, 2026
Search Tags:#calibration#forecasting#time-series#preflight#timesfm#covariates
One-click install
npx skills add https://github.com/JosephWoodall/noosphere --skill timesfm-forecasting-josephwoodall
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: timesfm-forecasting
Source: https://github.com/JosephWoodall/noosphere/tree/main/.agent/skills/timesfm-forecasting
Command: npx skills add https://github.com/JosephWoodall/noosphere --skill timesfm-forecasting-josephwoodall

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, pandas, torch, timesfm, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Unsupervised, training-free forecasting of univariate time series using Google's TimesFM foundation model, enabling accurate forecasts with calibrated prediction intervals without requiring model training.

Core Features & Use Cases

  • Zero-shot forecasting for univariate time series from CSV/DataFrame/arrays with quantile intervals.
  • Batch processing of multiple series and flexible context/horizon configuration via ForecastConfig.
  • Optional covariates support through forecast_with_covariates for timesfm[xreg].

Quick Start

Load TimesFM from Hugging Face, compile with a ForecastConfig, and forecast your first univariate time series.

Frequently Asked Questions about timesfm-forecasting

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I perform zero-shot time-series forecasting without training a model?▼

Zero-shot time-series forecasting is achievable using Google's TimesFM foundation model to deliver point forecasts without training. This approach uses a pre-trained model to generate predictions directly from your univariate data.

Can I generate calibrated quantile prediction intervals for univariate time series?▼

Yes, calibrated quantile prediction intervals are natively supported for univariate time series forecasting. The system delivers these intervals alongside point forecasts without requiring any model training.

How do I forecast multiple time series in batch from a CSV or DataFrame?▼

Batch processing of multiple time series is supported directly from CSV, DataFrame, or numpy arrays. You can configure context and horizon parameters via ForecastConfig to process many series efficiently.

Do I need a specific GPU or RAM setup to run TimesFM forecasting?▼

A mandatory preflight check verifies available RAM, GPU/VRAM, and disk space before loading models. This ensures safe and scalable deployment by preventing out-of-memory errors during the forecasting process.

Can I include external covariates in my time-series forecasts?▼

External covariates are supported through the forecast_with_covariates function for timesfm[xreg]. This allows you to incorporate additional variables to refine univariate time-series predictions.

What is the best way to configure context length and forecast horizon for time-series predictions?▼

Configuring context length and forecast horizon is handled through the ForecastConfig object. This flexible configuration allows you to tailor the zero-shot forecasting parameters to fit your specific univariate data structure.

Related Skills

Explore More Data & Analytics Skills→
vasilyu1983vasilyu1983
73

ai-ml-timeseries

Forecast time series data with LightGBM, Transformers, and SHAP.

Community
Advanced
swaruplabswaruplab
94

timesfm-forecasting

Forecast univariate time series with the TimesFM 2.5 model.

Community
Advanced
AlterLab-IEUAlterLab-IEU
58

alterlab-timesfm

Forecast univariate time-series data with zero-shot TimesFM point and quantile forecasts.

Community
Advanced
biomaps-infrabiomaps-infra

bio-timesfm-forecasting

Generates zero-shot time series forecasts with Google's TimesFM from CSV, DataFrame, or array inputs.

Official
Advanced