My dataset consists of weekly sales data of thousands of products, for hundreds of stores. Occasionally, there is a promotion in one week, leading to spikes in the sales of the promoted product. During the promotion week, the sales on other products can drop, i.e. cannibalization.
I am a little familiar with the arules package, which can help identify association rules between products.
Yet, how could we identify the products being cannibalized through the sales promotions? Is there a similar package like arules, just for cannibalization?