Box-Cox transformation on dataset from compositional studies of archaeological potteries

被引：3

作者：

Santos, Jose O. ^{[1
]}

Reis, Michael Santana ^{[1
]}

Munita, Casimiro S. ^{[2
]}

Silva, Jose E. ^{[1
]}

机构：

[1] Inst Fed Educ, Ciencia & Tecnol Sergipe IFS, Ave Lourival Baptista S-N, BR-79000000 Lagarto, SE, Brazil

[2] Inst Pesquisas Energet & Nucl IPEN CNEN SP, Ave Prof Lineu Prestes 2242,Cidade Univ, BR-05508000 Sao Paulo, Brazil

来源：

JOURNAL OF RADIOANALYTICAL AND NUCLEAR CHEMISTRY | 2017年 / 311卷 / 02期

关键词：

Geochemistry; Archaeometry; MVN test; Box-Cox transformation; INAA; MULTIVARIATE; ARCHAEOMETRY; NORMALITY; CERAMICS; OUTLIERS;

D O I：

10.1007/s10967-016-4987-y

中图分类号：

O65 [分析化学];

学科分类号：

070302 ; 081704 ;

摘要：

In archaeometric data set it was verified by statistical tests that some variables almost never follow a multivariate normal distribution using logarithmic transformation or other. This work presents a multivariate Box-Cox transformation for MardiaA ' s and RoytonA ' s tests for a data set of fifty ceramic fragments from archaeological site Justino, Xing, Brazil, and one clay sample collected near of the site. The samples were analyzed by instrumental neutron activation analysis, INAA. The study was made using companion to applied regression package from R software and was tested by Hotelling's T (2) statistics.

引用

页码：1427 / 1433

页数：7

共 44 条

[1] CORRECTIONS FOR DRILL-BIT CONTAMINATION IN SAMPLING ANCIENT-POTTERY FOR NEUTRON-ACTIVATION ANALYSIS [J].