Can we have multiple branches in the Paxata Data Prep Pipeline? Example- Can I split the Paxata Data prep pipeline into two based on filter conditions. I will call 2 different ML models in each sub pipeline.
The Predict tool supports the option to filter. So, if you have a column or series of columns based on which two different models need to be used, you could add the predict tool, apply the filter, make predictions. Repeat this process for the n segments in your data.
The entire prediction will remain in the Paxata Project as one dataset, you could then perform any post-prediction data prep followed by publishing the data out. Even for publishing out, you could apply the same filters on the Lenses so that you have multiple outputs.