i work with your library for an introductory course in Data Science. Problem is, that i can not get any valid results with the KNN-Prediction Algorithms. Every time i run these Algorithms the mean is calculated. I have created my own Dataset. Sorry to get in contact but i am desperate. With using SVD or SVDpp i get predictions for the users.
Regards,
Andreas
Steps/Code to Reproduce
reader = Reader(rating_scale=(0,100))
data = Dataset.load_from_df(df2[['bon',"hauptgruppe","anzahl"]].astype(str), reader)
trainset = data.build_full_trainset()
algo = KNNBasic()
algo.fit(trainset)
pred1=algo.predict(str(0),str(0))
print (pred1)
User and item are existing:
Expected Results
That the user and the item will be identified and a prediction can be made, like in the examples.
Description
Hello Nicolas,
i work with your library for an introductory course in Data Science. Problem is, that i can not get any valid results with the KNN-Prediction Algorithms. Every time i run these Algorithms the mean is calculated. I have created my own Dataset. Sorry to get in contact but i am desperate. With using SVD or SVDpp i get predictions for the users.
Regards, Andreas
Steps/Code to Reproduce
reader = Reader(rating_scale=(0,100)) data = Dataset.load_from_df(df2[['bon',"hauptgruppe","anzahl"]].astype(str), reader) trainset = data.build_full_trainset()
algo = KNNBasic() algo.fit(trainset)
pred1=algo.predict(str(0),str(0)) print (pred1)
User and item are existing:
![grafik](https://user-images.githubusercontent.com/89678913/150836015-0d04d424-a322-42c4-b20a-d4b5e66b56b3.png)
Expected Results
That the user and the item will be identified and a prediction can be made, like in the examples.
Actual Results
Versions