(Preprocessing + Feature Engineering) Place function to aggregate data

[GregW04]

In scope of this task we need figure out when to aggregate whole data. This is important because some functions already built may suffer if we put in the wrong place. For this to happen you need to contact whole team about already built code in whole Preprocessing and FE.

From my point of view it should be somewhere before Feature Engineering.

After finding potential spaces contact with @GregW04 with proposed solution and also proposed aggregation levels.