General description of the training project

I: General information of the training

Title: RooStats tutorial
Dates: 26 - 27 November 2009
Internet address of the learning environment:
Internet address of the agenda:
Last update: 24 Nov, 2009
Steering group: Lorenzo Moneta, Grégory Schott, Wouter Verkerke
Lecturers: see above
Tutors: see above + appr. 4-5 from ATLAS (Max Baak, Hella Snoek, Alfio Lazarro), CMS (Danilo Piparov, Mario Pelliccioni)
Local organization & technical support: Kati Lassila-Perini (CERN)
Teaching level: under-graduate, post-graduate, further education
Discipline: statistical data analysis (software development)
Estimated number of participants: 50 ATLAS+CMS
Main goal of the training:
- participants: learn to use statistical tools provided by Root/RooStats with statistical methods typically needed in CMS/ATLAS analysis
- project team: train (use the tools) and educate (understand the statistical concepts behind) future users, get feedback to improve the product and the documentation
Pedagogical approach of the training:
- Short introductory lecture
- Walk-through of a working example code explaining it row-by-row
- Hands-on exercise based on examples
Every participant will have an analysis project in mind and the outcome of the training should be applicable to the project (but not applied during the training).
Brief description of the scenario - tasks and expected results:
- basic knowledge of root, see e.g.
- basic knowledge of statistical methods, lectures by Luca Lista on the week before
Session 1 -
  • Introduction to RooStats project (Kyle - 30')
    • background of the project
    • general concepts of statistics
    • RooStats class architecture
  • Introduction to RooFit (Wouter - 30')
  • RooFit tutorials (hands-on session) (Wouter 1h)
    • exercise on how to create a model and save in a workspace, toy-MC, ...
Session 2 -
  • Introduction to some statistical methods available in RooStats (Gregory 30')
    • Detailed description of 3 basic statistical methods (likelihood, frequentist and bayesian)
    • Examples on how they are implemented and can be used in RooStats
    • Introduction to exercises
  • RooStats basics tutorial (hands-on session 45') (Gregory and Lorenzo)
    • Exercises on the 3 methods (likelihood, frequentist and bayesian). Significance and limits but without systematics with a simple model
    • Model: (Gaussian + flat background) and leave number counting as home-work exercise (or the other way around)
Session 3 -
  • RooStats tutorials (hands-on session 1) (1h30') (Gregory, Kyle, Lorenzo)
    • Continue previous exercises but with systematics
    • Other exercises with more complex models (to be defined) and let users play variations (who prepares that?)
  • Introduction to advanced methods in RooStats (30') (Kyle or Lorenzo)
    • Introduction to other tools to be used in the last session (NeymanConstruction, MCMC )
    • slides presenting tools not covered by hands-on sessions (e.g. sPlot, HLFactory ? )
  • RooStats tutorials (hands-on session 2) (1h00') (Gregory, Kyle, Lorenzo)

II: Organization

Ratio: Presence/remote 80%/20% (remote participation via EVO for the mini-lecture at the beginning of each session)
Brief description of the management of the training - tasks and roles:
Steering group: plan a learning material and provide contents
guarantee the coherence
Lecturers: present material and guide the exercises
update the existing material
Tutors: help with the exercises, reply to questions
Participants: complete the exercises, give feedback
Tools: web area for links to contents and for providing feedback, tarball for code (or link to cvs), videoconferencing for transmission and recording
Management tools: Web site, indico, code repository, e-mail (project team)

III: General roles and tasks

People Role Task
Steering group: Responsible of the pedagogical content and coherence during planning: intervene in the planning of the learning material by the content providers
Experts on separate learning modules Provide or update learning material
Lecturers: Teaching Give lectures
Tutors: Guide the participant in their exercises Assist to the hands-on sessions
Run the tutorials beforehand
Technical support: Practical arrangements Organize the rooms and EVO connections and recordings
Course-related technical support Help with technical issues connected to the course (availability of the material, accessibility)
Project management: Take care of the timely execution of the plan Plan and manage the project
Participants: Give feedback
Interaction between the protagonists:
- project team: direct contact, e-mail discussion
- discussion sessions and questions during the lectures
Tools for interaction:
- project description and organization documents (wiki), meetings, e-mail for the project team
Constraints and limits:
- timing (clashes with major CMS/ATLAS events)
- availability of lecturers and tutors
- internet connectivity
- prior knowledge of participants
- new root version and compatibility of the tutorials

IV: Evaluation modalities

Evaluation criteria of the different activities: - no formal evaluation
Feedback: - on training: logbook area for each session and a feedback form in the end of the course
- on the toolkit: discussion, logbook keeping track on the difficulties encountered.

-- KatiLassilaPerini - 27-Oct-2009

Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r3 - 2009-11-24 - KatiLassilaPerini
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Sandbox All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback