Hello, If working with a static partitioned asset if I have some partitions that are not materialized or even if I delete some partitions that were bad whilst waiting for new data is there any way I can use just the available partitions in a downstream asset. For example I have some monitors and I use them to train a model. I realize that one is miscalibrated so I delete that materialized partition but waiting for the recalibration process is too much downtime so I want to build a model on just the other partitions that are available. When I pass the partitioned asset into a new asset that would reduce across the partitions it errors when looking for the deleted partition. Can this type of flexibility be accommodated or do I have to shift to manually managing the partitions with

DynamicOutputs

?

hi @Rhys Goodall! I created a github discussion for this question with the answer below 🙂