Metrohm Vision Manual - Manual (Page 108)

106 ▪▪▪▪▪▪▪
4. You can edit the local identification method by clicking on Identify/Apply Math and following
the identification method development procedure.
5. Save the new method by clicking on Identify/Save Results.
6. Click on Identify/Return to Library Clustering when the method is edited and saved.
Saving the Clustering Method
The library clustering method is saved automatically. A new clustering method will overwrite the
existing one as soon as it is created.
Deleting the Clustering Method
The clustering method can be deleted by clicking on Clustering/Delete Cluster Tree. A clustering
method has to be deleted before a global identification method is created.
Adding new products to the library will invalidate the library clustering method. An appropriate
warning will be displayed in the Sample Selection program.
4.5 Identification Method Development
4.5.1 Introduction
Identification is a very important step in qualitative analysis. Vision handles identification in two
different ways, depending on whether a clustering method has been created for the library.
If the library does not have a clustering method created, a single identification method can be
developed. This method is global for the library in the sense that it applies to all the products in the
library with a single identification threshold. On the other hand, if a Principal Component based
identification method is used (Mahalanobis distance or residual variance), a local PC model for each
product is calculated.