BUCHI NIRCal - Manual (Page 183)

Chemometrics
NIRCal 5.5 Manual, Version A 183
3.18.42 Instrument
Description Contains the used instrument index.
Use For 1D-scatter and dependency plots. Can be used for Outlier selection.
Method PCR / PLS / Cluster (CLU) / SIMCA / MLR
Matrices ID 104
Tip
Details Instrument index is a column vector.
Related Topic
Description
Contains the used instrument index.
Use For 1D-scatter and dependency plots. Can be used for Outlier selection.
Method PCR / PLS / Cluster (CLU) / SIMCA / MLR
Matrices ID 105
Tip
Details Instrument index is a row vector.
Related Topic
3.18.43 Leverages
Description Mahalanobis distance from the center of the score space to each spectra.
Use To find Leverage Outliers.
Method PCR / PLS / Cluster (CLU) / SIMCA
Matrices ID 32
Tip
Details All primary PCs are used. The secondary PC selection has no effect on the
leverages.
Related Topic Scores
Description
Contains the Leverages against the PCs for the different properties (Cluster).
Use For PC selection.
Method PCR / PLS / Cluster (CLU) / SIMCA
Matrices ID 94
Tip
Details
Related Topic Leverages
3.18.44 Loadings or Principal Components
Description Loadings build up the base for reconstructing the spectra together with the
scores and the
Eigenvalues .
Use Look for the wavelengths activity in the loadings and compare it with the
original spectra.
Method PCR / PLS / Cluster (CLU) / SIMCA
Matrices ID 4
Tip Mark interesting bands in the wavelengths and have a look to the spectra.
The wavelengths selection will also appear in the plot.
Details Loadings are called also Principal Components or sometimes till called
"factors" (NIRCal 4 terminology). The maximum available number of PCs
calculated is called the primary PCs.
The selected PCs for predicting properties are called secondary PCs.
Related Topic Scores , Eigenvalues
The loadings are artificial difference spectra. Only principal components with characteristic spectral
information should be used.