Getting started#

fasthep-flow is typically used through the fasthep command-line interface and the broader FAST-HEP package ecosystem.

The recommended installation method is the fasthep meta package, which provides curated installation profiles for different use cases.

Warning

The FAST-HEP ecosystem is currently undergoing a major rewrite and alpha-stage reorganisation.

Not all packages are available on PyPI yet, and some installation details may change while the ecosystem stabilises.

We will try to keep the main installation entry points stable for their intended use cases.

Installation profiles#

The fasthep meta package provides several installation profiles.

Minimal installation#

Recommended for:

  • workflow language exploration

  • backend/runtime development

  • lightweight experimentation

  • non-HEP workflows

Installs:

  • fasthep-flow

  • fasthep-cli

pip install "fasthep[minimal]"

This is the smallest supported FAST-HEP installation profile without any HEP-specific dependencies.

HEP installation#

Recommended for most HEP users.

Includes:

  • workflow execution

  • ROOT and awkward support

  • transforms and histogramming

  • rendering

  • diagnostics

  • CLI tooling

pip install "fasthep[hep]"

This is expected to become the standard installation profile for typical HEP analyses.

Full installation#

Installs the complete FAST-HEP ecosystem and optional tooling.

pip install "fasthep[full]"

This profile is mainly intended for:

  • developers

  • integration testing

  • ecosystem experimentation

Verifying the installation#

After installation, verify that the CLI is available:

fasthep version

You can also inspect installed package versions:

fasthep versions

Next steps#

There are two common paths after installation.

Learn the workflow system#

If you want to understand how fasthep-flow works internally, continue with:

  1. Workflow language

  2. Compilation pipeline

  3. Execution model

These pages explain:

  • declarative workflows

  • dependency inference

  • execution planning

  • runtime orchestration

  • backend abstraction

  • serialisable execution plans

Start building workflows immediately#

If you want to start running workflows and tutorials right away, head to the FAST-HEP Workshop:

The workshop contains:

  • beginner tutorials

  • runnable example workflows

  • toy datasets

  • rendering examples

  • workflow debugging exercises

  • advanced backend and extensibility examples

The workshop is the recommended hands-on entry point for most users.

Development installations#

For active development, editable installs, and integration testing, see https://github.com/FAST-HEP/fasthep-dev