HANA PAL (double exponential smoothing) multiple data sets and speed

Problem -

I'm attempting to implement multiple stactical forecasts across several data sets. Currently in SQLSCRIPT I am looping at my data sets and calling out to DoubleSmooth implementation and returning the result, ideally what I would like to do is pass a single large data set 12 months for 12 different sets and get 12 different forecast results, is this possible ? should I be using a different PAL/BFL function to execute the forecast (that has this capability).

The intent of this is to;

1. Remove the loop that current marshals the data for the call into the PAL function.

2. Allow HANA to optimize / parallel the calls to increase speed of processing.

Any ideas would be greatly appreciated.