I can’t use the pointcloudcoercer and then group on ”count” since it still need to be in pointcloud- format when writing to Xyz-fileformat. Do I really need to first write to xyz and then bring it back in to split it up?
Thanks
Page 1 / 1
Hi @abbe11
I'd recommend using the PointCloundExpressionEvaluator to create a new component, the expression should be something like @index()/1000. Then, use a PointCloudSplitter to split by that component.
I've attached some images and a workspace as an example. 
Have you looked at the PointCloudDensityTiler on FME Hub?
https://hub.safe.com/publishers/safe-lab/transformers/pointclouddensitytiler
Thanks @jdh , that almost does what I am after. However, Is it also possible to get a "tile-id" to achieve the final split/fanout on?
Hi @abbe11
I'd recommend using the PointCloundExpressionEvaluator to create a new component, the expression should be something like @index()/1000. Then, use a PointCloudSplitter to split by that component.
I've attached some images and a workspace as an example.
Thanks @siennaatsafe, this works well. The newly created component "t" isn't exposed as an component, right? (It is still possible to use it in pointcloudsplitter though, but not as a component in "split by"- drop down.)
Thanks @jdh , that almost does what I am after. However, Is it also possible to get a "tile-id" to achieve the final split/fanout on?
You could just add a Counter to get an id for the fanout.
Thanks @siennaatsafe, this works well. The newly created component "t" isn't exposed as an component, right? (It is still possible to use it in pointcloudsplitter though, but not as a component in "split by"- drop down.)
Hi @abbe11 ,
That's right, it won't count as an exposed component, but can be written in the pointcloudsplitter (however not selected by the dropdown).