Jarvis-HEP: A lightweight Python framework for workflow composition and parameter scans in high-energy physics

TL;DR

Jarvis-HEP is a lightweight Python framework designed to simplify workflow composition and parameter scans in high-energy physics, integrating multiple tools and supporting flexible, dependency-aware execution.

Contribution

It introduces a YAML-based, dependency-aware workflow system with modular components and built-in sampling backends tailored for high-energy physics phenomenology.

Findings

Enables complex multi-step computational studies in high-energy physics.

Supports both external software and internal components within a unified workflow.

Includes built-in sampling backends for exploratory parameter scans.

Abstract

High-energy physics phenomenology often requires linking multiple computational tools to evaluate observables, likelihoods, and experimental constraints across nontrivial parameter spaces. In this work, we introduce Jarvis-HEP, a lightweight Python framework for workflow composition and parameter scans in high-energy physics. The framework provides YAML-based workflow specification, dependency-aware execution, modular calculator integration, and asynchronous task scheduling for multi-step computational studies. It supports both external software packages and internally implemented components within a unified workflow, and the current implementation includes several built-in sampling backends for exploratory scans. This paper describes the design and user interface of Jarvis-HEP and illustrates its use with representative synthetic and phenomenological examples.