SUM(IF(FREQUENCY(IF(ISNUMBER(MATCH($A$1:$G$5000,N44:O44,0)),ROW($A$1:$G$5000)),ROW($A$1:$G$5000))=2,1))

Hello Stackoverflowers,

I have the following problem: SUM(IF(FREQUENCY(IF(ISNUMBER(MATCH($A$1:$G$5000,N44:N45,0)),ROW($A$1:$G$5000)),ROW($A$1:$G$5000))=2,1)) This formula lists any 2 items in an extensive purchasing list of customers. I would like to narrow it down somehow for a few weeks now, and can't get close enough with it. Would like to be able to do a better inventory prediction. That is, if for example (unsure if my screenshot goes through)

banana cucumber apple chocolate vitamins detergent meat cheese cake candy gas pork

lemon banana orange cookie cucumber bread sweets comics ice cream fork lamb

soap make up lemon towel banana tickets vitamins news chocolate fudge spoon

orange lip stick banana flour cucumber fabric softener gas lemon bread cookie bbq

banana diesel soda candles biscuit apple knife cake lemon spoon tonic

If a customer buys banana, cucumber, cookie and lemon, we can list those, and see what items, products are the most popular with those articles they are buying. For example, I know if it is rainy, more tonic water is needed, but what articles they buy the most with tonic...? The idea would be if we could list 3-15 items, row by row together, depending on the list of the receipt, and know by the purchased items, if birthdays are coming up, or any celebrations in the neighborhood by seeing statistical history, when they bought certain items the most to order for the future to come. No Cambridge Analytica, just PO predictions.

Thank you very much!