Question

What are Feature Counters?

Answer

Feature Counters are the (optional) display of the number of features that have passed through a connection within an FME workspace. This functionality is available in FME2006-GB or newer.

The display of feature counts occurs after a translation has been executed. They help a user to analyze the results of a workspace, and provide a reference for debugging if the destination data differs from what was expected.



Above: The 'Display Feature Counts' button appears on the FME Workbench toolbar.



Above: In this workspace 22 features were read, processed and written.


However, not all links always show the same number, especially when using transformers that affect the number of features (for example the DuplicateRemover or 2DPointReplacer transformer).



Above: Here 84 features entered a DuplicateRemover transformer but only 11 emerged, enabling us to calculate that there were 73 duplicate features.


Feature Counts also help illustrate the nature of how data can be split, duplicated or merged.



Above: This DuplicateRemover transformer shows how it automatically splits 84 features into two streams of 11 and 73.



Above: Here a manual splitting takes place. By drawing two output connections an extra copy of data is created - 22 features in each stream (instead of dividing the 22 features into 2 streams of 11).


So splitting the data creates a new copy - but what about bringing two streams together? That will add the two streams together, rather than trying to merge them.



Above: Here a user is mistakenly trying to merge 10 spatial features with their 10 related database records. Unfortunately, FME doesn't work this way and the result is a straightforward 20 features (10 spatial, 10 non-spatial).



Above: This user has mastered the trick of merging data - the key is to use a FeatureMerger transformer. 10 spatial features and 10 database records = 10 spatial features with attached attributes.