13 Statistical Techniques for the Interpretation of Analytical Data 703
Table 13.16Results of applying LDA to the data of 10 volatile compounds in order to discriminate
the four groups of varietal wines. Classification functions:
Malvar Air ́en Trepat Monastrell
Methanol 10.799 10.257 0.256 5.117
1-Propanol 22.060 19.429 –10.130 4.492
Isobutanol 7.578 10.969 9.830 3.557
Isoamylic alcohols 2.758 2.107 –2.919 0.675
1-Hexanol –426.708 –326.405 357.505 –62.965
cis-3-Hexen-1-ol –689.479 –641.758 442.618 –86.874
Hexanoic acid 18.915 19.051 5.176 11.013
Octanoic acid –25.394 –27.983 11.366 –8.608
Decanoic acid 7.380 40.821 153.799 61.847
Ethyl octanoate –32.454 –7.918 81.870 6.050
Constant –818.041 –813.798 –761.745 –259.392
Table 13.17Results of applying LDA to the data of 10 volatile compounds in order to discriminate
the four groups of varietal wines. Posterior probabilities:
Wine:
Observed
classif. Malvar Air ́en Trepat Monastrell
Ma98 Malvar 1.00 0.00 0.00 0.00
Ma98 Malvar 1.00 0.00 0.00 0.00
Ma99 Malvar 1.00 0.00 0.00 0.00
Ma99 Malvar 1.00 0.00 0.00 0.00
Ai97 Air ́en 0.00 1.00 0.00 0.00
Ai97 Air ́en 0.00 1.00 0.00 0.00
Ai98 Air ́en 0.00 1.00 0.00 0.00
Ai98 Air ́en 0.00 1.00 0.00 0.00
Tr97 Trepat 0.00 0.00 1.00 0.00
Tr97 Trepat 0.00 0.00 1.00 0.00
Tr98 Trepat 0.00 0.00 1.00 0.00
Tr98 Trepat 0.00 0.00 1.00 0.00
Mo97 Monastrell 0.00 0.00 0.00 1.00
Mo97 Monastrell 0.00 0.00 0.00 1.00
Mo98 Monastrell 0.00 0.00 0.00 1.00
Mo98 Monastrell 0.00 0.00 0.00 1.00
Table 13.18Results of applying LDA to the data of 10 volatile compounds in order to discriminate
the four groups of varietal wines. Classification matrix:
Predicted classifications:
Observed Percent
classifications correct Malvar Air ́en Trepat Monastrell
Malvar 100.0 4 0 0 0
Air ́en 100.0 0 4 0 0
Trepat 100.0 0 0 4 0
Monastrell 100.0 0 0 0 4
Total 100.0 4 4 4 4