You need JavaScript to view this

On the use of the covariance matrix to fit correlated data

Abstract

The best fits to data which are affected by systematic uncertainties on the normalization factor have the tendency to produce curves lower than expected, if the covariance matrix of the data points is used in the definition of the {chi}{sup 2}. This paper shows that the effect is a direct consequence of the hypothesis used to estimate the empirical covariance matrix, namely the linearization on which the usual error propagation rely. The bias can become unacceptable of the normalization error is large, or a large number of data points are fitted. (orig.)
Authors:
D`Agostini, G [1] 
  1. Rome Univ. 2 (Italy). Dipt. di Fisica
Publication Date:
Dec 01, 1993
Product Type:
Technical Report
Report Number:
DESY-93-175
Reference Number:
SCA: 990200; PA: DE-94:0G7194; NTS-94:018480; EDB-94:086025; ERA-19:020532; SN: 94001207896
Resource Relation:
Other Information: PBD: Dec 1993
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; LEAST SQUARE FIT; CORRELATIONS; MATRICES; NONLINEAR PROBLEMS; ERRORS; DATA PROCESSING; 990200; MATHEMATICS AND COMPUTERS
OSTI ID:
10155256
Research Organizations:
Deutsches Elektronen-Synchrotron (DESY), Hamburg (Germany)
Country of Origin:
Germany
Language:
English
Other Identifying Numbers:
Journal ID: ISSN 0418-9833; Other: ON: DE94769009; TRN: DE94G7194
Availability:
OSTI; NTIS (US Sales Only)
Submitting Site:
DE
Size:
9 p.
Announcement Date:
Jul 06, 2005

Citation Formats

D`Agostini, G. On the use of the covariance matrix to fit correlated data. Germany: N. p., 1993. Web.
D`Agostini, G. On the use of the covariance matrix to fit correlated data. Germany.
D`Agostini, G. 1993. "On the use of the covariance matrix to fit correlated data." Germany.
@misc{etde_10155256,
title = {On the use of the covariance matrix to fit correlated data}
author = {D`Agostini, G}
abstractNote = {The best fits to data which are affected by systematic uncertainties on the normalization factor have the tendency to produce curves lower than expected, if the covariance matrix of the data points is used in the definition of the {chi}{sup 2}. This paper shows that the effect is a direct consequence of the hypothesis used to estimate the empirical covariance matrix, namely the linearization on which the usual error propagation rely. The bias can become unacceptable of the normalization error is large, or a large number of data points are fitted. (orig.)}
place = {Germany}
year = {1993}
month = {Dec}
}