The maxRecommendations parameter is used when you want less recommendations than the model generates.
By default, the Frequently Bought Together (FBT) model recommends up to 3 products, so using a maxRecommendations bigger than 3 won’t have an effect.
You can change the number of recommendations from 3 to 30, and after the model is re-trained, you will see up to 4 recommendations (if the maxRecommendations parameter is set to 4)