Uncertainty Quantification UQ Project - Quality Investigation and Software Metrics Analysis

Quantitative investigation of physics foundations and observables for simulation codes, collaboration with research institutes, analysis of Geant4 software quality, and ongoing/future activities in data libraries and workshops.

• Uncertainty
• Quantification
• UQ Project
• Software Metrics
• Geant4

cfere Follow

Uploaded on Feb 18, 2025 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. Uncertainty Quantification (UQ) Project CCR-funded project 2012-2019 Elisabetta Ronchieri as member of UQ team INFN CNAF, 20 December 2019

2. Objectives O1 Quantitative investigation of the quality of physics foundations and observables for simulation codes and applications It includes tests for the validation of physics models and observables used in Monte Carlo particle transport code, e.g. Geant4, Fluka, MCNP, and so on And mathematical framework for uncertainty quantification O2 Quantitative investigation of the quality of software that produces physics foundation and observables High quality software is necessary for high quality physics Objective assessment of the quality of Geant4 Geant4 is a playground for the development of a methodology for the quality assessment of physics software

3. UQ Team UQ team changed over the project duration Close collaboration with Max Planck Institute in Munich (G. Waidenspointener), Hanyang University in Seul (C. H. Kim), PUC University of Porto Alegre (G. Hoff), IXFEL in Hamburg (M. Kuster) Table shows INFN team members in 2019. Objectives People N. FTE O1, O2 Maria Grazia Pia 1 INFN Genoa O1 Paolo Saracco 0.4 O1 Min Cheol 1 INFN CNAF O2 Elisabetta Ronchieri 0.2

4. Results: O2 Results: O2 Issues Results TO DO or In Progress Existing software metrics data sets are old, not always well documented, not related to scientific domains, not free, Built a new set of software metrics data sets, specific for Geant4: - 34 releases examined; - for each release built data set at file, class, variable, and function; - metrics are related to complexity, software size and adhesion to object orientation paradigm. 2. Make data sets available to software engineering communities and Geant4 users Existing software metrics analysis are based on these data sets: they are usually empirical studies Performed quantitative analysis by using statistical methods: - trend analysis to monitor software evolution; - inequality analysis to determine how metrics distribution change as a software system evolves 1. A journal article is under review (to be submitted at Empirical Software Engineering: An International Journal https://www.springer.com/journal/10664) Some software metrics tools are old, not for C++ languages, include couple of metrics Selected Imagix 4D. Its collaboration supports research activities (always obtained full free license both for windows and linux) 3. Request a new full license Geant4 team @ CERN have: showed interest in results, given support to obtain code and provided feedback to understand some results.

5. Ongoing and Future activities Definition of a new project, IDataLib on physics data libraries (leader Genova) Workshop Open Data Libraries Workshop at IEEE NSS MIC 2019 https://nssmic.ieee.org/2019/workshops/ A Manifesto document is under definition (M. G. Pia @ INFN, E. Ronchieri @ INFN, M. Fleming @ International Nuclear Data Evaluation Co-operation, C. Hill Head of IAEA Atomic and Molecular Data Unit, D. Brown Head of ENDF/B @ Brookhaven National Laboratory, B. Prytichenko @ Brookhaven National Laboratory, K. Tada @ Japan Atomic Energy Agency, R. Townson Research Officer for National Research Council Canada, M.-A. Descalle @ Xray Science & Technology Group Leader at Lawrence Livermore National Laboratory, M. Spannowsky Prof. Centre for Particle Theory @ Durham University, T. Basaglia Head of CERN Library @ CERN Scientific Information Service Application of Unsupervised and Semi-supervised ML techniques on software characteristics data sets (CNAF) Up to now 4 articles on international conferences 1 article on journal is under submission Journal of Systems and Software https://www.journals.elsevier.com/journal-of-systems-and-software