The principle of PCA classification and Fuzzy Clustering (FC) is the reduction of the dimensionality of a spectral space to a level comprehensible to the human eye. Since a paper or screen plot is inherently 2D, the method of choice is usually 2D PCA or FC. However, if a 3D projection on a 2D computer screen is accompanied with motion, the user can perceive this simulation as 3D space. To increase the space feeling the 3D plots are interactive, i.e. the plot can be rotated with the mouse. Spectra Projector allows interactive 3D viewing of your data.

The 3D projection mode used in Mass Frontier stems from the idea that spectra classification of complex data set in a 3D space can be more effective and reliable than a simple 2D plot. This idea is based on the well-recognized fact that the human brain is incredibly efficient in visual pattern recognition in 3D space and can still outperform any computer system.

When analyzing PCA or FC results you should not rely entirely on 2D plots. Two or more clusters which overlap in a 2D plot may be separated in 3D space or, on the other hand, two seemingly separate clusters discernable in a 2D plot may appear in 3D mode to be too close together to be considered as two different objects.

To view PCA or FC results in 3D mode:

1. Click the 3D Projection button .
2. Click anywhere in the classification plot with the left mouse button and move the mouse to rotate the plot interactively. Note that the arrow cursor will be replaced by a circle during the dragging step.

To better appreciate the three-dimensionality of the space you can display the x-, y-, and z- axes by clicking the Show Axes button , which is next to the 3D Projection button.

Note: In the 3D projection mode it is NOT possible to enlarge a part of the plot. However, as in the 2D mode, the selection of particular spectra and the display of their origins is supported.