How do I feed multiple inputs to an HDA through the HDA processor in PDG?

Hi there,

There is a section in the parameters for specifying the inputs to the HDA. When the input source is “Upstream Output Files”, all the input files with the specified tag will be used as inputs to the HDA. If you select “Custom File Paths”, you can manually specify the inputs to the HDA.

Hi Brandon

Thanks for the reply. But if my HDA is looking for 3 specific inputs for eg. How do I control what get's fed into each input? The HDA processor only allows for one input. Does it use the File Tag to somehow determine multiple inputs? If so, can you give an example?

Thanks!

To elaborate.. one of the HDA inputs takes geometry, the other one takes txt files. How do I feed this into the HDA processor?

With “Upstream Output Files”: the inputs are wired according to the ordering of the outputs in the upstream node. So the first output will become the first input to the HDA, the second output will be wired as the HDA's second input, etc. If you have inputs that you want to provide to the HDA Processor that are produced by multiple nodes, you will need to partition together those work items first before feeding it to the HDA Processor (and ensure that ‘Merge Input Attributes’ is enabled on the partitioner).

With the “Custom File Paths” option: the inputs are wired according to their numbering (Input File 1 -> 1st input pin of the HDA, Input File 2 -> 2nd input pin of the HDA, etc).

Doesn't each input act as it's own work item? When I feed a list of objects into the HDA processor it runs the HDA per obj. It doesn't wire each object in as a file input to the same HDA and then run them all together. So how does it know to treat the inputs as file inputs to the same HDA, rather than running the HDA per input?

HDA Processor will generate 1 work item per upstream work item. If you want to provide multiple file inputs from different upstream nodes to HDA Processor, you will first need to partition those work items together before the HDA Processor node. I've attached a .hip file that demonstrates this. From the Tasks menu select “Generate Static Work Items” to see how it works (cooking it won't work as I haven't hooked up an HDA to the HDA Processor – this is just for demonstration).

If you attach your .hip file I can also take a look at it for you.

I left something out of the hip file: you will also want to turn on “Create File Inputs” on the HDA Processor node and change the File Tag parameter to just “file”.

Thanks for this Brandon!