I am fairly new to HANA PAL and this is what I have been trying out.
The scenario I have is the following: -
I need to calculate forecast for multiple materials based on their sales history. I have the data organized in my table as follows: -
| Material | Calendar Month | Sales Qty |
|---|---|---|
| ABC | Jan-15 | 200 |
| ABC | Feb-15 | 400 |
| ABC | Mar-15 | 125 |
| ABC | Apr-15 | 300 |
| XYZ | Jan-15 | 50 |
| XYZ | Feb-15 | 20 |
| XYZ | Mar-15 | 30 |
| XYZ | Apr-15 | 10 |
... and so on
My final output needs to look like the following
| Material | Forecast Qty |
|---|---|
| ABC | 400 |
| XYZ | 25 |
I have already created the generated procedure using my PAL function however the limitation is that I would need to invoke the procedure for Material ABC (with four rows) just to calculate the forecast for it. Then once again for material XYZ, and continue the operation for every material.
The way I understood this is that it because in the "input data table" parameter I can only send one time series data at a time. I may have 100,000 materials for which I need to execute forecast.
How can I accomplish this without having to invoke my procedure 100,000 times in a LOOP? Is there a better way?
By the way, this is a typical problem that I noticed for practically any PAL function.