UnROOT.jl

UnROOT.jl is a reader for the CERN ROOT file format written entirely in Julia, without any dependence on ROOT or Python.

Important API changes in v0.9.0

Click to expand example for RNTuple

We decided to alter the behaviour of getindex(f::ROOTfile, s::AbstractString) which is essentially the method called called when f["foo/bar"] is used. Before v0.9.0, UnROOT tried to do a best guess and return a tree/branch or even fully parsed data. This lead to two bigger issues.

Errors prevented any further exploration once UnROOT bumped into something it could not interpret, although it might not even be requested by the user (e.g. the interpretation of a single branch in a tree, while others would work fine)
Unpredictable behaviour (type instability): the path dictates which type of data is returned.

Starting from v0.9.0 we introduce an interface where f["..."] always returns genuine ROOT datatypes (or custom ones if you provide interpretations) and only performs the actual parsing when explicitly requested by the user via helper methods like LazyBranch(f, "...").

Long story short, the following pattern can be used to fix your code when upgrading to v0.9.0:

f("foo/bar") => LazyBranch(f, "foo/bar")

The f["foo/bar"] accessor should now work on almost all files and is a handy utility to explore the ROOT data structures.

See PR199 for more details.

Installation Guide

Download the latest Julia release
Open up Julia REPL (hit ] once to enter Pkg mode, hit backspace to exit it)

julia>]
(v1.8) pkg> add UnROOT

Quick Start (see docs for more)

TTree

julia> using UnROOT

julia> f = ROOTFile("test/samples/NanoAODv5_sample.root")
ROOTFile with 2 entries and 21 streamers.
test/samples/NanoAODv5_sample.root
├─ Events (TTree)
│  ├─ "run"
│  ├─ "luminosityBlock"
│  ├─ "event"
│  ├─ "⋮"
│  ├─ "L1_UnpairedBunchBptxPlus"
│  ├─ "L1_ZeroBias"
│  └─ "L1_ZeroBias_copy"
└─ untagged (TObjString)


julia> mytree = LazyTree(f, "Events", ["Electron_dxy", "nMuon", r"Muon_(pt|eta)$"])
 Row │ Electron_dxy                      nMuon   Muon_pt          Muon_eta        
     │ SubArray{Float3                   UInt32  SubArray{Float3  SubArray{Float3 
─────┼────────────────────────────────────────────────────────────────────────────
 1   │ [0.000371]                        0       []               []
 2   │ [-0.00982]                        2       [19.9, 15.3]     [0.53, 0.229]
 3   │ []                                0       []               []
 4   │ [-0.00157]                        0       []               []
 5   │ []                                0       []               []
 6   │ [-0.00126]                        0       []               []
 7   │ [0.0612, 0.000642]                2       [22.2, 4.43]     [-1.13, 1.98]
 8   │ [0.00587, 0.000549, -0.00617]     0       []               []
  ⋮  │                ⋮                    ⋮            ⋮                ⋮
                                                                  992 rows omitted

RNTuple

Click to expand example for RNTuple

julia> using UnROOT

julia> f = ROOTFile("./test/samples/RNTuple/test_ntuple_stl_containers.root");

julia> f["ntuple"]
UnROOT.RNTuple with 5 rows, 13 fields, and metadata:
  header: 
    name: "ntuple"
    ntuple_description: ""
    writer_identifier: "ROOT v6.29/01"
    schema: 
      RNTupleSchema with 13 top fields
      ├─ :lorentz_vector ⇒ Struct
      ├─ :vector_tuple_int32_string ⇒ Vector
      ├─ :string ⇒ String
      ├─ :vector_string ⇒ Vector
      ├─ :vector_vector_int32 ⇒ Vector
      ├─ :vector_variant_int64_string ⇒ Vector
      ├─ :vector_vector_string ⇒ Vector
      ├─ :variant_int32_string ⇒ Union
      ├─ :array_float ⇒ StdArray{3}
      ├─ :tuple_int32_string ⇒ Struct
      ├─ :array_lv ⇒ StdArray{3}
      ├─ :pair_int32_string ⇒ Struct
      └─ :vector_int32 ⇒ Vector
      
  footer: 
    cluster_summaries: UnROOT.ClusterSummary[ClusterSummary(num_first_entry=0, num_entries=5)]

julia> LazyTree(f, "ntuple")
 Row │ string  vector_int32     array_float      vector_vector_i     vector_string       vector_vector_s     variant_int32_s  vector_variant_     ⋯
     │ String  Vector{Int32}    StaticArraysCor  Vector{Vector{I     Vector{String}      Vector{Vector{S     Union{Int32, St  Vector{Union{In     ⋯
─────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
 1   │ one     [1]              [1.0, 1.0, 1.0]  Vector{Int32}[Int3  ["one"]             [["one"]]           1                Union{Int64, Strin  ⋯
 2   │ two     [1, 2]           [2.0, 2.0, 2.0]  Vector{Int32}[Int3  ["one", "two"]      [["one"], ["two"]]  two              Union{Int64, Strin  ⋯
 3   │ three   [1, 2, 3]        [3.0, 3.0, 3.0]  Vector{Int32}[Int3  ["one", "two", "th  [["one"], ["two"],  three            Union{Int64, Strin  ⋯
 4   │ four    [1, 2, 3, 4]     [4.0, 4.0, 4.0]  Vector{Int32}[Int3  ["one", "two", "th  [["one"], ["two"],  4                Union{Int64, Strin  ⋯
 5   │ five    [1, 2, 3, 4, 5]  [5.0, 5.0, 5.0]  Vector{Int32}[Int3  ["one", "two", "th  [["one"], ["two"],  5                Union{Int64, Strin  ⋯
                                                                                                                                  5 columns omitted

LazyTree as unified table / iteration interface

You can iterate through a LazyTree:

julia> for event in mytree
           @show event.Electron_dxy
           break
       end
event.Electron_dxy = Float32[0.00037050247]

julia> Threads.@threads :static for event in mytree # multi-threading
           ...
       end

Only one basket per branch will be cached so you don't have to worry about running out of RAM. At the same time, event inside the for-loop is not materialized until a field is accessed. This means you should avoid double-access, see performance tips

XRootD is also supported, depending on the protocol:

the "url" has to start with http:// or https://:
(1.6+ only) or the "url" has to start with root:// and have another // to separate server and file path

julia> r = ROOTFile("https://scikit-hep.org/uproot3/examples/Zmumu.root")
ROOTFile with 1 entry and 18 streamers.
https://scikit-hep.org/uproot3/examples/Zmumu.root
└─ events (TTree)
   ├─ "Type"
   ├─ "Run"
   ├─ "Event"
   ├─ "⋮"
   ├─ "phi2"
   ├─ "Q2"
   └─ "M"

julia> r = ROOTFile("root://eospublic.cern.ch//eos/root-eos/cms_opendata_2012_nanoaod/Run2012B_DoubleMuParked.root")
ROOTFile with 1 entry and 19 streamers.
root://eospublic.cern.ch//eos/root-eos/cms_opendata_2012_nanoaod/Run2012B_DoubleMuParked.root
└─ Events (TTree)
   ├─ "run"
   ├─ "luminosityBlock"
   ├─ "event"
   ├─ "⋮"
   ├─ "Electron_dxyErr"
   ├─ "Electron_dz"
   └─ "Electron_dzErr"

TBranch of custom struct

We provide an experimental interface for hooking up UnROOT with your custom types that only takes 2 steps, as explained in the docs. As a show case for this functionality, the TLorentzVector support in UnROOT is implemented with the said plug-in system.

Support & Contributiing

Use Github issues for any bug reporting or feature request; feel free to make PRs, bug fixing, feature tuning, quality of life, docs, examples etc.
See CONTRIBUTING.md for more information and recommended workflows in contributing to this package.

Acknowledgements

Special thanks to Jim Pivarski (@jpivarski) from the Scikit-HEP project, who is the main author of uproot, a native Python library to read and write ROOT files, which was and is a great source of inspiration and information for reverse engineering the ROOT binary structures.

Contributors ✨

Thanks goes to these wonderful people (emoji key):

_{Tamas Gal} 💻 📖 🚇 🔣 ⚠️	_{Jerry Ling} 💻 ⚠️ 🔣 📖	_{Johannes Schumann} 💻 ⚠️ 🔣	_{Nick Amin} 💻 ⚠️ 🔣	_{Mosè Giordano} 🚇	_{Oliver Schulz} 🤔	_{Misha Mikhasenko} 🔣
_{Yuan-Ru Lin} ⚠️