XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1461
of misclassification is lesser for the calcic and magnesian
skarns, although the latter two also comprise a volumetric
lesser number of samples. However, the suggested discrimi-
nation biplot is able to classify correctly about 85% of the
samples. and thus, it is considered of good quality for the
purpose given the significant, lithological, and mineralogi-
cal variability at copper deposits. This classification diagram
has the potential to be used widely in other similar mineral-
izing environments to characterize the spatial distribution
of skarn related alteration based on geochemistry.
Sulphide Associations
A representative subset of 271 samples with combined
mineralogy as whole rock chemistry were used for DFA to
prediction of the dominant sulphides/oxide associations,
following a protocol similar to that outlined earlier for the
skarns. Four groups of samples were a-priory determined
based on the mineralogical data using PCA and balances
(Figure 5 a,b). The graph illustrates that the data along the
x axis account for most of the variability and distinguish
the sulphides dominant mineralogy toward positive val-
ues, while the negative values define mineral assemblages
associated dominantly with oxide phases. Values along
the y axis further separate pyrite- and bornite-dominated
assemblages.
As expected, when the data plotted on a binary S vs Cu
diagram, the groups defined by the TIMA-X mineralogy
display a good correlation with the expected stoichiometric
Figure 5. PCA (centred log ratios) and balances used for the a-priory classification of sulfide groups
Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1461
of misclassification is lesser for the calcic and magnesian
skarns, although the latter two also comprise a volumetric
lesser number of samples. However, the suggested discrimi-
nation biplot is able to classify correctly about 85% of the
samples. and thus, it is considered of good quality for the
purpose given the significant, lithological, and mineralogi-
cal variability at copper deposits. This classification diagram
has the potential to be used widely in other similar mineral-
izing environments to characterize the spatial distribution
of skarn related alteration based on geochemistry.
Sulphide Associations
A representative subset of 271 samples with combined
mineralogy as whole rock chemistry were used for DFA to
prediction of the dominant sulphides/oxide associations,
following a protocol similar to that outlined earlier for the
skarns. Four groups of samples were a-priory determined
based on the mineralogical data using PCA and balances
(Figure 5 a,b). The graph illustrates that the data along the
x axis account for most of the variability and distinguish
the sulphides dominant mineralogy toward positive val-
ues, while the negative values define mineral assemblages
associated dominantly with oxide phases. Values along
the y axis further separate pyrite- and bornite-dominated
assemblages.
As expected, when the data plotted on a binary S vs Cu
diagram, the groups defined by the TIMA-X mineralogy
display a good correlation with the expected stoichiometric
Figure 5. PCA (centred log ratios) and balances used for the a-priory classification of sulfide groups

Help

loading