Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Geochemical Discrimination Diagrams: A Simple Way of Replacing Eye-Fitted Field Boundaries With, Probability Based Classifier Surfaces


Affiliations
1 Department of Geology, University of Rajasthan, Jaipur 302 004, India
     

   Subscribe/Renew Journal


The subjective, eye-fitted, field boundaries in geochemical discrimination diagrams constitute one of the possible reasons for their often-noted inability to determine the source regime/tectonic environments of igneous rocks. A simple method is described here to replace the eye-fitted field boundaries between the groups with the probability based classifier surfaces. It involves projection of the assignment rules obtained from linear discriminant analysis onto the binary or ternary discrimination diagram of raw variables. The method is illustrated by obtaining classifier surface between I- and S-type granites in Na2O-K2O space and between Anorthosite/Rapakivi (AR) and Ring Complex (RC) granites in Na2O-MgO space. The need for explicitly ensuring adequacy of samples, selection of suitable variables, evaluation of rate of misclassification and a better understanding of the unique influence of tectonic environments on magmatic process is emphasized for optimizing the performance of the classifier surface.

Keywords

Igneous Rocks, Geochemical Discrimination, Tectonic Classification, I- and S-type Granitoids, Rapakivi and Ring-Complex Granites.
Subscription Login to verify subscription
User
Notifications
Font Size

Abstract Views: 243

PDF Views: 2




  • Geochemical Discrimination Diagrams: A Simple Way of Replacing Eye-Fitted Field Boundaries With, Probability Based Classifier Surfaces

Abstract Views: 243  |  PDF Views: 2

Authors

Salil Agrawal
Department of Geology, University of Rajasthan, Jaipur 302 004, India

Abstract


The subjective, eye-fitted, field boundaries in geochemical discrimination diagrams constitute one of the possible reasons for their often-noted inability to determine the source regime/tectonic environments of igneous rocks. A simple method is described here to replace the eye-fitted field boundaries between the groups with the probability based classifier surfaces. It involves projection of the assignment rules obtained from linear discriminant analysis onto the binary or ternary discrimination diagram of raw variables. The method is illustrated by obtaining classifier surface between I- and S-type granites in Na2O-K2O space and between Anorthosite/Rapakivi (AR) and Ring Complex (RC) granites in Na2O-MgO space. The need for explicitly ensuring adequacy of samples, selection of suitable variables, evaluation of rate of misclassification and a better understanding of the unique influence of tectonic environments on magmatic process is emphasized for optimizing the performance of the classifier surface.

Keywords


Igneous Rocks, Geochemical Discrimination, Tectonic Classification, I- and S-type Granitoids, Rapakivi and Ring-Complex Granites.