At Novonesis, we use spectroscopy (mainly IR and NIR) for investigating complex matrices. The data are commonly explored and analyzed using principal components analysis (PCA) or partial least squares regression (PLS-R), which are methods that respect and utilize the correlation between the individual wavelengths/wavenumbers. These correlations open up for enhanced data exploration and model building but also require proper visualization tools in software to ensure correct model fit and to present model outcomes to scientists, managers, and other stakeholders. 

In this presentation, we show how we use PCA and PLS at Novonesis to gain a better understanding of the complex matrices; we also highlight some missing features in PCA and PLS in JMP.