Analysis of counting data: Development of the SATLAS Python package

For the analysis of low-statistics counting experiments, a traditional nonlinear least squares minimization routine may not always provide correct parameter and uncertainty estimates due to the assumptions inherent in the algorithm(s). In response to this, a user-friendly Python package (SATLAS) was...

Full description

Saved in:
Bibliographic Details
Published in:Computer physics communications Vol. 222; pp. 286 - 294
Main Authors: Gins, W., de Groote, R.P., Bissell, M.L., Granados Buitrago, C., Ferrer, R., Lynch, K.M., Neyens, G., Sels, S.
Format: Journal Article
Language:English
Published: Elsevier B.V 01.01.2018
Subjects:
Chi-square
Counting experiment
Data analysis
Likelihood
Python 3
Python 3
Counting experiment
Data analysis
Chi-square
Likelihood
ISSN:0010-4655, 1879-2944
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:For the analysis of low-statistics counting experiments, a traditional nonlinear least squares minimization routine may not always provide correct parameter and uncertainty estimates due to the assumptions inherent in the algorithm(s). In response to this, a user-friendly Python package (SATLAS) was written to provide an easy interface between the data and a variety of minimization algorithms which are suited for analyzinglow, as well as high, statistics data. The advantage of this package is that it allows the user to define their own model function and then compare different minimization routines to determine the optimal parameter values and their respective (correlated) errors. Experimental validation of the different approaches in the package is done through analysis of hyperfine structure data of 203Fr gathered by the CRIS experiment at ISOLDE, CERN. Source code: https://github.com/woutergins/satlas/ Documentation: https://woutergins.github.io/satlas/ Program Title: SATLAS Program Files doi:http://dx.doi.org/10.17632/3hr8f5nkhb.1 Licensing provisions: MIT Programming language: Python External routines/libraries: NumPy, SciPy, LMFIT, Pandas, NumDiffTools Nature of problem: Fitting data from a counting experiment to extract parameter information. Solution method: Supply a modular library with fitting routines using pre-implemented goodness-of-fit statistics for counting data under different circumstances.
ISSN:0010-4655
1879-2944
DOI:10.1016/j.cpc.2017.09.012