Wine Samples Analyzed by GC-MS and FT-IR

Different analytical instruments can describe complementary aspects of wines. This collection pairs headspace GC-MS and FT-IR measurements to compare variation shared between the techniques with variation unique to each one. The collection contains 44 red wines made from 100% Cabernet Sauvignon grapes and bought in Copenhagen: 6 from Argentina, 15 from Chile, 12 from Australia, and 11 from South Africa. Duplicate measurements were made for each wine, and the released blocks contain their averages. Wine_v7.mat contains Data_GC and Data_GC_unaligned, each a 44 x 2700 x 200 GC-MS array ordered sample x elution time x mass channel. The scans cover m/z 5-204. The file also contains mass-summed Elution_profiles, elution-time-summed Mass_profiles, and Aroma_compounds, peak areas for 57 labelled aroma compounds. The FT-IR blocks are IR_spectra (44 x 842, water bands removed) and IR_spectra_with_waterband (44 x 1056). Pred_values_IR contains 14 WineScan-predicted composition variables, including ethanol, acids, pH, reducing sugar, density, glycerol, methanol, and tartaric acid. Origin classes, sample labels, compound names, and the GC-MS and FT-IR axes are included. Original dataset page: Wine samples analyzed by GC-MS and FT-IR instruments Original ERDA package: Wine GC-MS FT-IR If you use these data, please cite:Skov, T.; Ballabio, D.; Bro, R. (2008). Multiblock variance partitioning: A new approach for comparing variation in multiple data blocks. Analytica Chimica Acta, 615(1), 18-29. https://doi.org/10.1016/j.aca.2008.03.045