I have two sets of sales data. One for year 2016. And next set of data for 2017 which I receive every week.
So assuming I am at first week of 2017. I so far have data only for week = 1.
- The items have some property like: category & brand; which are usually common in both year.
- However, the item (item_code) from this year (2017) won't be same as item from 2016.
- Sales quantity has been normalized by sum of total sales for that item.
Now that I get data for each week gradually for year 2017 what would be best possible way to normalize sales quantity?
Random idea occurred to me is to use features of item (brand, category) to some how approximate but I don't have concrete idea on how to proceed. Especially on R, to which I am quite new. So any suggestion with quick guide of how to on R would be appreciated.