Skip to main content

Hi

does anybody know of a good alternative to the spatial filter? Due to the volume of data involved it doesn't seem feasible to use the spatial filter but I can't really see an alternative. Unfortunately I have around 3 million polygons to pass through a filter of around 1 million polygons. I have used a feature reader to break up the data into manageable chunks, which makes the workspace run at a reasonable speed but this requires me to rerun it multiple times with each chunk. I was hoping to be able to run it in one go. Using a workspace runner has also helped with processing, but again, still not feasible to run it in one go.

Assuming you look for an alternative for the "SpatialFilter" transformer.

Not sure if this fits your case but what I often do is use a FeatureReader to read the "filter" polygons, then use another FeatureReader, using it's spatial filter, to read the "candidate" polygons. This way you don't have to compare all filters to all candidates.

Another powerful function is to merge the attributes from the filters with the candidates when being read. This way you can assign group id's to the candidates by which you can group by further downstream.


thanks for your reply. I managed to do it using a SQL Creator in the end, which worked well, sending batches at a time


thanks for your reply. I managed to do it using a SQL Creator in the end, which worked well, sending batches at a time

Hi,

I am facing the same issue in SpatialFilter. How did you use SQLCreator. Can you elaborate it?

Thanks

Kumar


Reply