Focus on:  Hide Contributions
Back to Conference View
Login

PyHEP 2019 Workshop

→ Europe/London
Abingdon, U.K.

Abingdon, U.K.

The Cosener's House Abingdon, U.K.
Benjamin Krikler (University of Bristol (GB)), Eduardo Rodrigues (University of Cincinnati (US))
Description

The PyHEP workshops are a series of workshops initiated and supported by the HEP Software Foundation (HSF) with the aim to provide an environment to discuss and promote the usage of Python in the HEP community at large. Further information is given here.

PyHEP 2019 will be held at The Cosener's House, in Abingdon, near Oxford, United Kingdom, from 16-18 October 2019.

The workshop will be a forum for the participants and the community at large to discuss developments of Python packages and tools, exchange experiences, and steer where the community needs and wants to go. There will be ample time for discussion.
 
The agenda will be composed of plenary sessions, a highlight of which is the following:
1) A keynote presentation on the PyViz - open source visualization tools for Python - project, given by Philipp Rudiger, a member of the developers team.
2) Topical sessions on e.g. histogramming and statistics, including a talk and a hands-on tutorial.
3) Lightning talks from participants.
4) Presentations following up from topics discussed at PyHEP 2018.
 
Partial travel support for some U.S. participants (in particular, students and early-career postdocs) may be available from the IRIS-HEP institute. Please contact Peter Elmer (Peter.Elmer@cern.ch) to enquire about details.

 

Organising Committee

Eduardo Rodrigues - University of Cincinnati (Chair)
Ben Krikler - University of Bristol (Co-chair)

 

Sponsors

The event is kindly sponsored by

          Python Software Foundation   STFC PPGP

Live Notes
PyHEP2019_Participants.jpg
PyHEP2019_Participants_thumbnail.png
workshop_photos.zip
Contact us
Participants
55 View full list
  • Wednesday 16 October
  • Thursday 17 October
    • 09:00 → 10:10
      Visualisation
      Convener: Benjamin Krikler (University of Bristol (GB))
      • 09:00
        Panel: Turn your existing Python analysis code into deployable dashboards 40m

        KEYNOTE PRESENTATION:

        Over the last decade the Python scientific ecosystem have become an incredibly powerful toolkit for performing analyses and visualizing data. Once an analysis is done it often has to be shared with a wider audience, either within an organization or with the wider public, but this step often requires an entirely different set of tools and skillset. Panel is a new, open-source Python library built to easily wrap the outputs of an analysis, combine them with widgets and then lay them out as an interactive app or dashboard. This enables faster iteration cycles within organizations and ensures users without in-depth familiarity with web programming can develop and deploy complex dashboards with minimal code.

        Panel natively supports a wide range of plotting tools and many other types of data making it trivial to work with the tools users are already familiar with. A complex interactive Panel-based dashboard is typically many times shorter than the equivalent Dash or Bokeh code, focusing on expressing relationships between widgets, computation, and visualizations directly. At the same time Panel is not limited to building simple apps and can be used to visualize complex, multi-stage analysis pipelines, provides full styling flexibility using CSS and Bokeh themes and has the ability to dynamically resize to the size of the browser window.

        Once an application is built either in a Python script or notebook it can trivially be deployed as a standalone app using Bokeh Server without any change in behavior. Alternatively, Panel apps can be exported to static HTML files by defining Javascript based interactions or even recording and embedding the app's state space, making it possible to share interactive visualizations as self-contained files.

        In the talk we will discover some of the core ideas behind Panel, go through the process of making an existing Jupyter notebook deployable as a dashboard and finally look at a number of case studies of real world dashboards analyzing large volumes of scientific data. With Panel, your analyses and visualizations can now very easily leap from your notebook or custom scripts into the real world!

        Speaker: Philipp Rudiger (Anaconda)
        Presentation repository
        PyHEP2019_slides.pdf
      • 09:40
        mpl-hep - HEP needs for visualization 30m
        Speaker: Andrzej Novak (RWTH Aachen (DE))
        PyHEP_2019_Andrzej.pdf
    • 10:10 → 10:40
      Coffee/tea break 30m
    • 10:40 → 11:40
      Analysis platforms
      Convener: Eduardo Rodrigues (University of Cincinnati (US))
    • 11:50 → 12:50
      High-level Analysis Tools
      Convener: Eduardo Rodrigues (University of Cincinnati (US))
    • 12:50 → 14:30
      Lunch 1h 40m
    • 14:30 → 15:30
      Analysis fundamentals
      Convener: Benjamin Krikler (University of Bristol (GB))
      • 14:30
        Writing files with uproot 30m
        Speaker: Pratyush Das (Institute of Engineering and Management, Kolkata)
        uproot-pyhep.pdf
      • 15:00
        Awkward 1.0 30m
        Speaker: Jim Pivarski (Princeton University)
        pivarski-pyhep-awkward.pdf
    • 15:30 → 16:00
      Coffee/tea break 30m
    • 16:00 → 18:00
      Histogramming & news on Python 3.8
      Convener: Henry Fredrick Schreiner (Princeton University)
      • 16:00
        Aside: quick news on Python 3.8 15m
        Speaker: Henry Fredrick Schreiner (Princeton University)
        python38.pdf
      • 16:15
        Python histogramming packages 45m
        Speaker: Henry Fredrick Schreiner (Princeton University)
        histogramming.pdf
      • 17:00
        Hands-on 1h
        Speaker: Henry Fredrick Schreiner (Princeton University)
        Workbook repo
    • 19:00 → 20:00
      Dinner 1h
  • Friday 18 October
    • 09:00 → 10:30
      Statistics
      Convener: Hans Peter Dembinski (Max-Planck-Institute for Nuclear Physics, Heidelberg)
      • 09:00
        zfit 30m
        Speaker: Jonas Eschle (Universitaet Zuerich (CH))
        pyhep_zfit_Jonas_Eschle.pdf
        zfit github
        zfit tutorials

        zfit: scalable pythonic fitting

        zfit brings together the model fitting efforts for HEP in Python by defining an API & Workflow and an implementation on top of TensorFlow and scikit-hep packages.

        It offers an extensive model building part, currently focused on unbinned likelihoods.

      • 09:30
        pyhf 30m
        Speaker: Matthew Feickert (Univ. Illinois at Urbana Champaign (US))
        Feickert_pyhf-talk-export.pdf
        GitHub repo hosting talk
        pyhf talk
      • 10:00
        Fast likelihood analysis in more dimensions for Xenon TPCs 30m
        Speaker: Bart Pelssers (Stockholm University)
        pyhep_talk.pdf
    • 10:30 → 11:00
      Coffee/tea break 30m
    • 11:00 → 12:30
      Statistics
      Convener: Eduardo Rodrigues (University of Cincinnati (US))
      • 11:00
        Introduction to iminuit 30m
        Speaker: Dr Hans Peter Dembinski (Max-Planck-Institute for Nuclear Physics, Heidelberg)
        iminuit_intro.html
      • 11:30
        Statistical Methods in the NPStat Package 30m
        Speaker: Igor Volobouev (Texas Tech University (US))
        install_npstat_with_prerequisites.csh
        npstat.pdf
      • 12:00
        Computing coverage with NumPy and Numba of Bayesian and Frequentist intervals 30m
        Speaker: Hans Peter Dembinski (Max-Planck-Institute for Nuclear Physics, Heidelberg)
        dembinski_coverage_numba.html
    • 12:30 → 13:00
      PyROOT
      Convener: Benjamin Krikler (University of Bristol (GB))
    • 13:00 → 13:05
      Closeout
    • 13:05 → 14:05
      Lunch 1h
Powered by Indico v3.3.6-pre