Premium
Outlier Treatment in Data Merging
Author(s) -
Blessing R. H.
Publication year - 1997
Publication title -
journal of applied crystallography
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.429
H-Index - 162
ISSN - 1600-5767
DOI - 10.1107/s0021889896014628
Subject(s) - outlier , weighting , statistics , mathematics , sample (material) , identification (biology) , anomaly detection , pattern recognition (psychology) , computer science , data mining , artificial intelligence , medicine , physics , radiology , botany , biology , thermodynamics
Experience with a variety of diffraction data‐reduction problems has led to several strategies for dealing with mismeasured outliers in multiply measured data sets. Key features of the schemes employed currently include outlier identification based on the values y median = median(| F i | 2 ), σ median = median[ σ (| F i | 2 )], and | Δ | median = median(| Δ i |) = median[|| F i | 2 ‐median (| F i | 2 )|] in samples with i = 1, 2 n and n ≥ 2 measurements; and robust/resistant averaging weights based on values of | z i | = | Δ i |/max{ σ median , | Δ | median [ n /( n −1)] 1/2 }. For outlier discrimination or down‐weighting, sample median values have the advantage of being much less outlier‐based than sample mean values would be.