Garnet major-element composition as an indicator of host-rock type: a machine learning approach using the random forest classifier / supplementary data

The database includes 13615 garnet compositions of eight oxides commonly analysed in lab routines: SiO₂, TiO₂, Al₂O₃, Cr₂O₃, FeO_total, MnO, MgO, and CaO (in wt%). These are complemented by the following covariables: <strong>setting and metamorphic facies class</strong>: code indicating the geologic/tectonic setting of the host rock <strong>composition class</strong>: code indicating the compositional class of the host rock <strong>author</strong>: authors of the original paper providing the data <strong>journal</strong>: journal of the original paper <strong>region</strong>: origin of the data, in the format "region, country" <strong>sample name</strong>: sample ID in the original paper <strong>Pavg(kbar)</strong>: if available, indicated pressure <strong>Tavg(°C)</strong>: if available, indicated temperature <strong>host-rock type and/or metamorphic facies</strong>: facies indication of host rock <strong>lithology and/or protolith</strong>: composition indication of host rock <strong>SiO2</strong>: wt% <strong>TiO2</strong>: wt% <strong>Al2O3</strong>: wt% <strong>Cr2O3</strong>: wt% <strong>FeOtotal</strong>: wt% <strong>MnO</strong>: wt% <strong>MgO</strong>: wt% <strong>CaO</strong>: wt% This research was funded by DFG grant EY 23/27-1.