Recommendations are coming same before and after fit_partial for new interactions for any user lightfm

Question

for any user let's say some new interactions come then I am fitting the dataset and model with only new interactions data using partial_fit and I am getting the same results before and after partial fit for that user.

Answer 1

I dont really understand the program but I think you are not getting the dessired results because you are not defining functions and hence you are working with global variables instead of local ones

try:

def makeitems(df2,user):
     ...
     return itemfeatures

I wanted to comment but I need 50 rep