def recommend_k_fastai(model, test, train):
with Timer() as t:
total_users, total_items = model.data.train_ds.x.classes.values()
total_items = total_items[1:]
total_users = total_users[1:]
test_users = test[DEFAULT_USER_COL].unique()
test_users = np.intersect1d(test_users, total_users)
users_items = cartesian_product(test_users, total_items)
users_items = pd.DataFrame(
users_items, columns=[DEFAULT_USER_COL, DEFAULT_ITEM_COL])
training_removed = pd.merge(
users_items,
train.astype(str),
on=[DEFAULT_USER_COL, DEFAULT_ITEM_COL],
how="left",
)
training_removed = training_removed[
training_removed[DEFAULT_RATING_COL].isna()][[
DEFAULT_USER_COL, DEFAULT_ITEM_COL
]]
topk_scores = score(
model,
test_df=training_removed,
user_col=DEFAULT_USER_COL,
item_col=DEFAULT_ITEM_COL,
prediction_col=DEFAULT_PREDICTION_COL,
top_k=DEFAULT_K,
)
return topk_scores, t
def predict_k(self, dataset: RecommendationDataset, k: int) -> pd.DataFrame:
total_users = dataset.data[dataset.user_col].unique()
total_items = dataset.data[dataset.item_col].unique()
users_items = cartesian_product(total_users, total_items)
users_items = pd.DataFrame(users_items, columns=[dataset.user_col, dataset.item_col])
training_removed = pd.merge(users_items, dataset.data,
on=[dataset.user_col, dataset.item_col], how='left')
training_removed = training_removed[training_removed[dataset.score_col].isna()][[dataset.user_col, dataset.item_col]]
return self.predict_scores(dataset.wrap_data(training_removed))
def recommend_k_fastai(model, test, train):
with Timer() as t:
total_users, total_items = model.data.train_ds.x.classes.values()
total_items = total_items[1:]
total_users = total_users[1:]
test_users = test[DEFAULT_USER_COL].unique()
test_users = np.intersect1d(test_users, total_users)
users_items = cartesian_product(test_users, total_items)
users_items = pd.DataFrame(users_items, columns=[DEFAULT_USER_COL, DEFAULT_ITEM_COL])
training_removed = pd.merge(users_items, train.astype(str), on=[DEFAULT_USER_COL, DEFAULT_ITEM_COL], how='left')
training_removed = training_removed[training_removed[DEFAULT_RATING_COL].isna()][[DEFAULT_USER_COL, DEFAULT_ITEM_COL]]
topk_scores = score(model,
test_df=training_removed,
user_col=DEFAULT_USER_COL,
item_col=DEFAULT_ITEM_COL,
prediction_col=DEFAULT_PREDICTION_COL,
top_k=DEFAULT_K)
return topk_scores, t
learn.export('export.pkl')
learner = load_learner(path="/Users/rohittikle/Documents/GitHub/recommenders/")
#st.write(learner)
total_users, total_items = learner.data.train_ds.x.classes.values()
total_items = total_items[1:]
total_users = total_users[1:]
#st.write("Items:",total_items,"Users:",total_users)
#st.write("Users:",total_users)
test_users = data_test[USER].unique()
test_users = np.intersect1d(test_users, total_users)
#st.write("Test Users:",test_users)
users_items = cartesian_product(np.array(test_users),np.array(total_items))
users_items = pd.DataFrame(users_items, columns=[USER,ITEM])
# st.write("Users:",users_items)
training_removed = pd.merge(users_items, data_train.astype(str), on=[USER, ITEM], how='left')
training_removed = training_removed[training_removed['Rating'].isna()][[USER,ITEM]]
# st.write("Training:",training_removed)
start_time = time.time()
# st.write("Start Time:",start_time)
top_k_scores = score(learner,
test_df=training_removed,
user_col=USER,
item_col=ITEM,
prediction_col=PREDICTION)
#st.write("TOP K SCORES:",top_k_scores)
test_time = time.time() - start_time