uproot

Uproot is a Python library for reading ROOT data files, a format widely used in high-energy physics to store large event datasets. It enables analysts to access data stored in TTrees without requiring a local ROOT installation, facilitating integration with the Python data ecosystem such as NumPy, SciPy, and Pandas. Uproot reads TTrees and their branches and returns data as NumPy arrays or as awkward arrays to handle jagged or nested structures.

Key features include support for local and remote ROOT files, lazy loading of data to minimize memory usage, and automatic conversion of ROOT data types to Python-native types. It can handle common ROOT payloads, including arrays and vectors, and is designed to work with irregular data shapes. Uproot is part of the Scikit-HEP ecosystem and often used in conjunction with awkward-array for irregular data and with Coffea for scalable analyses.

Typical workflows involve opening a ROOT file with uproot.open, selecting a TTree, and requesting a subset of branches via arrays or similar methods, producing data structures ready for analysis, plotting, or further processing in Python. Because uproot is implemented in pure Python, it does not require the ROOT C++ libraries to be installed, and it can access data from local or remote sources.

Limitations and scope: uproot focuses on reading ROOT files and does not aim to be a complete ROOT I/O replacement. Some advanced or highly specialized ROOT payloads or user-defined classes may require additional handling or conversion. Writing ROOT files is not a primary feature of uproot.

Development and licensing: uproot is an open-source project maintained by contributors within the Scikit-HEP community, with its source code hosted on public repositories. It is widely used to enable Python-based data analysis pipelines in contemporary high-energy physics research.