Installation & Setup¶

This page describes how to set up the analysis environment on a local machine or a remote cluster.

System & Environment¶

Python ≥ 3.10

Required for all analysis scripts and notebooks.
Git

For cloning the repository.
C++ Compiler

Required for XRootD and compiled dependencies. GCC or Clang on Linux/macOS; use Conda on Windows.
Conda (recommended)

Handles complex binary dependencies reliably. Requires Miniconda or Anaconda. pip + virtualenv also supported.

Installation

All dependencies are installed automatically via environment.yml (Conda) or requirements.txt (pip). See Set Up the Python Environment for instructions.

1. Clone the Repository¶

Terminal

git clone https://github.com/anrghv/H-to-WW-NanoAOD-analysis.git
cd H-to-WW-NanoAOD-analysis

2. Set Up the Python Environment¶

Conda (recommended) pip (virtual environment)

The repository includes a complete environment.yml specifying all required packages with minimum version constraints:

Create and activate the environment

conda env create -f environment.yml
conda activate HEP_analysis

This creates a Conda environment named HEP_analysis with:

All Scikit-HEP packages (uproot, awkward, vector, hist)
Dask for distributed computing
JupyterLab for interactive notebooks
fsspec-xrootd for XRootD file access

Create and activate the virtual environment

python3 -m venv .venv
source .venv/bin/activate   # Linux / macOS
# .venv\Scripts\activate    # Windows
pip install -r requirements.txt

Windows

The analysis runs on Windows, macOS, and Linux. On Windows, Conda is strongly recommended -- some dependencies have complex build requirements that Conda resolves automatically.

3. Verify the Installation¶

Package check

Verify all packages

    import uproot, awkward as ak, vector, hist, dask
    print("All packages loaded successfully.")
    print(f"  uproot  : {uproot.__version__}")
    print(f"  awkward : {ak.__version__}")
    print(f"  dask    : {dask.__version__}")

XRootD connectivity

Test CERN EOS access

    import fsspec
    with fsspec.open(
        "root://eospublic.cern.ch//eos/opendata/cms/mc/"
        "RunIISummer20UL16NanoAODv9/GluGluHToWWTo2L2N_M-125"
        "_TuneCP5_minloHJJ_13TeV-powheg-jhugen727-pythia8/"
        "NANOAODSIM/106X_mcRun2_asymptotic_v17-v2/30000/"
        "00B3B6E3-3D68-C048-A8C4-04EB699CCE5D.root"
    ) as f:
        print("XRootD connection OK:", f.path)

4. Run the Analysis¶

Launch the notebook

cd notebooks/
jupyter lab HWW_analysis.ipynb

For a step-by-step walkthrough and instructions for running a full batch job with Dask, see the Analysis notebook.