Hi @maciejkula.
First: This library is amazing
Second: I would like to know how to add a user to the model. It's easy to add the user to the interaction matrix:
interactions.loc[user_id] = 0
But do I need to fit the model with all the interactions or only the new one:
# this
x = sparse.csr_matrix(interactions.loc[user].values)
# or this
x = sparse.csr_matrix(interactions)
model.fit(interactions=x, epochs=epoch, num_threads=n_jobs)
Hi @maciejkula. First: This library is amazing Second: I would like to know how to add a user to the model. It's easy to add the user to the interaction matrix:
But do I need to fit the model with all the interactions or only the new one: