atlasopenmagic 1.3.0

A utility package for retrieving ATLAS open data URLs and metadata.

Atlas Open Magic 🪄📊

Atlas Open Magic is a Python package made to simplify working with ATLAS Open Data by providing utilities to manage metadata and URLs for streaming the data.

Installation

You can install this package using pip.

pip install atlasopenmagic

Alternatively, clone the repository and install locally:

git clone https://github.com/atlas-outreach-data-tools/atlasopenmagic.git
cd atlasopenmagic
pip install .

Quick start

First, import the package:

import atlasopenmagic as atom

See the available releases and set to one of the options given by available_releases()

atom.available_releases()
set_release('2024r-pp')

Check in the Monte Carlo Metadata which datasets do you want to retrieve and use the 'Dataset ID'. For example, to get the metadata from Pythia8EvtGen_A14MSTW2008LO_Zprime_NoInt_ee_SSM3000:

all_metadata = atom.get_metadata('301204')

If we only want a specific variable:

xsec = atom.get_metadata('301204', 'cross_section')

To get the URLs to stream the files for that MC dataset:

all_mc = atom.get_urls('301204')

To get some data instead, check the available options:

atom.available_data()

And get the URLs for the one that's to be used:

all_mc = atom.get_urls('2016')

Open Data functions description and usage

available_releases()

Shows the available open data releases keys and descriptions.

Usage:

import atlasopenmagic as atom
atom.available_releases()

get_current_release()

Retrieves the release that the package is currently set at.

Usage:

release = atom.get_current_release()
print(release)

set_release(release, local_path)

Set the release (scope) in which to look for information (research open data, education 8 TeV, et). The release passed to the function has to be one of the keys listed by available_releases(). If a release has been downloaded locally, it is possible to use local_path to point atlasopenmagic towards the location of the datasets to use. Files are expected to be in a single directory. Metadata will be still retrieved remotely.

Additionally, if your facility mounts or mirrors EOS public as a local filesystem, you can provide "eos" as the local_path. In this case, files will be accessed using the native POSIX path

Args:

• release: name of the release to use.
• local_path: local directory path to use for downloaded datasets. If provided, the client will assume that datasets are available locally at this path. If 'eos' is provided, the original folder structure for the paths is kept.

Usage (remote storage):

atom.set_release('2024r-pp')

Usage (local storage):

atom.set_release('2024r-pp', local_path='/myproject/opendata/2024r-pp')

get_all_info(key, var)

Retrieves all the information for a given dataset, identified by its number or physics short name.

Args:

• key: The dataset identifier (e.g., '301204').
• var: A specific metadata field to retrieve. If None, the entire metadata dictionary is returned.

Usage: You can get a dictionary with all the metadata

metadata = atom.get_metadata('301209')

Or a single variable

xsec = atom.get_metadata('301209', 'cross_section')

The available variables are: dataset_number, physics_short, e_tag, cross_section_pb, genFiltEff, kFactor, nEvents, sumOfWeights, sumOfWeightsSquared, process, generator, keywords, file_list, description, job_path, CoMEnergy, GenEvents, GenTune, PDF, Release, Filters, release, skims.

The keys to be used for research data are the Dataset IDs found in the Monte Carlo Metadata

find_all_files(local_path, warnmissing=False)

Replace cached remote URLs in _metadata with corresponding local file paths if those files exist in the given local_path.

This function only affects the currently active release, and requires _metadata to be populated (it will trigger a fetch automatically).

Args:

• local_path (str): Root directory of your local dataset copy. Can have any internal subdirectory structure; only filenames are used for matching.

• warnmissing (bool, optional, default=False): If True, issue a UserWarning for every file that is in metadata but not found locally.

get_metadata(key, var)

Retrieves the metadata (no file lists) for a given MC dataset.

Args:

• key: The dataset identifier (e.g., '301204').
• var: A specific metadata field to retrieve. If None, the entire metadata dictionary is returned.

Usage: You can get a dictionary with all the metadata

metadata = atom.get_metadata('301209')

Or a single variable

xsec = atom.get_metadata('301209', 'cross_section')

The available variables are: dataset_number, physics_short, e_tag, cross_section_pb, genFiltEff, kFactor, nEvents, sumOfWeights, sumOfWeightsSquared, process, generator, keywords, description, job_path, CoMEnergy, GenEvents, GenTune, PDF, Release, Filters, release.

The keys to be used for research data are the Dataset IDs found in the Monte Carlo Metadata

get_urls(key, skim, protocol, cache)

Retrieves the list of URLs corresponding to a given key.

Args:

• key: Dataset ID.
• skim: Skim for the dataset. This parameter is only taken into account when using the 2025e-13tev-beta release.
• protocol: protocol for the URLs. Options: 'root' and 'https'.
• cache: use the simplecache mechanism of fsspec to locally cache files instead of streaming them.

Usage:

urls = atom.get_urls('12345', protocol='root', cache=True)

available_data()

Retrieves the list of keys for the data available for a scope/release.

Usage:

atom.available_data()

❗DEPRECATED get_urls_data(data_key, protocol)

Please use get_urls(key, skim='noskim', protocol=protocol, cache=False) instead.

Retrieves the list of URLs corresponding to one of the keys listed by available_data().

Args:

• data_key : For non-beta releases (e.g. '2015', '2016', etc.), the data key to look up.
• skim : Only for the 2025e-13tev-beta release: the skim name to look up.

Usage:

data = get_urls_data(data_key='2016', protocol='https')

Notebooks utilities description and usage

install_from_environment(*packages, environment_file)

Install specific packages listed in an environment.yml file via pip.

Args:

• *packages: Package names to install (e.g., 'coffea', 'dask').
• environment_file: Path to the environment.yml file. If None, defaults to the environment file for the educational resources.

Usage:

import atlasopenmagic as atom
atom.install_from_environment("coffea", "pandas", environment_file="./myfile.yml")

build_dataset(samples_defs, skim='noskim', protocol='https', cache=False)

Build a dict of data and / or MC samples URLs.

Args:

• samples_defs: Dictionary with DIDs and optional color: { sample_name: {'list': [...urls...], 'color': ...}, … }
• skim : The MC skim tag (only meaningful in the 2025e-13tev-beta release)
• protocol : Protocol to use for URLs.
• cache: use the simplecache mechanism of fsspec to locally cache files instead of streaming them.

Usage:

import atlasopenmagic as atom
atom.set_release('2025e-13tev-beta')
samples_defs = {
    r'Data':                    {'dids': ["data"],                      'color': 'red'},
    r'Background $t\bar t$':    {'dids': [410470],                      'color': 'yellow'},
    r'Background $V+$jets':     {'dids': [700335,700336,700337],        'color': 'orange'},
    r'Background Diboson':      {'dids': [700488,700489,700490,700491],'color': 'green'},
    r'Background $ZZ^{*}$':     {'dids': [700600,700601],               'color': '#ff0000'},
    r'Signal ($m_H$=125 GeV)':  {'dids': [345060,346228],              'color': '#00cdff'},
}

mc_samples = atom.build_dataset(samples_defs, skim='2bjets', protocol='https', cache=True)

Contributing

Contributions are welcome! To contribute:

1. Fork the repository.
2. Create a new branch (git checkout -b feature-name).
3. Commit your changes (git commit -am 'Add some feature').
4. Push to the branch (git push origin feature-name).
5. Create a Pull Request.

Please ensure all tests pass before submitting a pull request (just run pytest from the main directory of the package).

License

This project is licensed under the Apache 2.0 License