+1Hi folks,
I've been scratching my head with this one a little.
I've got features each with an appended list of values (features that are within X distance). I would like to cycle through the the features and get the average value of matches stored (I have 7) within the list.
The list for each feature is not a fixed size, 0-50.
Any suggestions about what the best approach would be?
+7Hi,
1. Make sure every feature does have a unique id
2. Use ListExploder to create features for every list element
(you might want to take a look at 'Conflict Resolution')
3. Use StatisticsCalculator to calculate average value grouped by id
You can get the list average using a ListSummer and a ListElementCounter, followed by an ExpressionEvaluator with something like
@Value(_sum)/@double(@Value(_element_count))This avoids using the ListExploder, which has a tendency to consume a lot of memory and slow down your translation considerably if you have many features and/or list elements. It also leaves you with the same number of output features and doesn't require a unique ID.
+7You can get the list average using a ListSummer and a ListElementCounter, followed by an ExpressionEvaluator with something like
@Value(_sum)/@double(@Value(_element_count))This avoids using the ListExploder, which has a tendency to consume a lot of memory and slow down your translation considerably if you have many features and/or list elements. It also leaves you with the same number of output features and doesn't require a unique ID.
+62Alternatively I think you can avoid exploding (an rebuilding) the list like this:
+62You can get the list average using a ListSummer and a ListElementCounter, followed by an ExpressionEvaluator with something like
@Value(_sum)/@double(@Value(_element_count))This avoids using the ListExploder, which has a tendency to consume a lot of memory and slow down your translation considerably if you have many features and/or list elements. It also leaves you with the same number of output features and doesn't require a unique ID.
+62
+1You can get the list average using a ListSummer and a ListElementCounter, followed by an ExpressionEvaluator with something like
@Value(_sum)/@double(@Value(_element_count))This avoids using the ListExploder, which has a tendency to consume a lot of memory and slow down your translation considerably if you have many features and/or list elements. It also leaves you with the same number of output features and doesn't require a unique ID.
I got to say, FME Knowledge Center is a great resource. I always get multiple good answers, really quickly too!
Thanks for all the help folks!