dislib.recommendation.ALS

class dislib.recommendation.als.base.ALS(random_state=None, n_f=100, lambda_=0.065, tol=0.0001, max_iter=100, arity=5, check_convergence=True, verbose=False)

Bases: sklearn.base.BaseEstimator

Alternating Least Squares recommendation.

Implements distributed alternating least squares recommendation based on Zhou et al. [1].

Parameters:

• max_iter (int, optional (default=100)) – Maximum number of iterations to perform.
• tol (float, optional (default=1e-4)) – Tolerance for the stopping criterion.
• n_f (int, optional (default=100)) – Number of latent factors (i.e. dimensions) for the matrices U and I.
• lambda_ (float, optional (default=0.065)) – Regularization parameters value.
• check_convergence (boolean, optional (default=True)) – Whether to test for convergence at the end of each iteration.
• random_state (int, orNone, optional (default=None)) – The seed of the pseudo random number generator used to initialize the items matrix I.
• arity (int, optional (default=5)) – The arity of the tasks during the merge of each matrix chunk.
• verbose (boolean, optional (default=False)) – Whether to print progress information.

Variables:

• users (np.array) – User matrix.
• items (np.array) – Items matrix.
• converged (boolean) – Whether the model has converged.

References

[1]	Zhou Y., Wilkinson D., Schreiber R., Pan R. (2008) Large-Scale Parallel Collaborative Filtering for the Netflix Prize. In: Fleischer R., Xu J. (eds) Algorithmic Aspects in Information and Management. AAIM 2008. Lecture Notes in Computer Science, vol 5034. Springer, Berlin, Heidelberg

Examples

>>> import dislib as ds
>>> from dislib.recommendation import ALS
>>> import numpy as np
>>> from scipy.sparse import csr_matrix
>>>
>>>
>>> if __name__ == '__main__':
>>>     data = np.array([[0, 0, 5], [3, 0, 5], [3, 1, 2]])
>>>     ratings = csr_matrix(data).transpose().tocsr()
>>>     train = ds.array(ratings, block_size=(1, 3))
>>>     als = ALS()
>>>     als.fit(train)
>>>     print('Ratings for user 0: %s' % als.predict_user(user_id=0))

fit(x, test=None)

Fits a model using training data. Training data is also used to check for convergence unless test data is provided.

Parameters:

• x (ds-array, shape=(n_ratings, n_users)) – ds-array where each row is the collection of ratings given by a user
• test (csr_matrix) – Sparse matrix used to check convergence with users as rows and items as columns. If not passed, uses training data to check convergence.

load_model(filepath, load_format='json')

Loads a model from a file. The model is reinstantiated in the exact same state in which it was saved, without any of the code used for model definition or fitting. :Parameters: * filepath (str) – Path of the saved the model

• load_format (str, optional (default=’json’)) – Format used to load the model.

Examples

>>> from dislib.recommendation import ALS
>>> import numpy as np
>>> import dislib as ds
>>> als2 = ALS()
>>> als2.load_model("model_als")
>>> predictions2 = als2.predict_user(user_id=0)

predict_user(user_id)

Returns the expected ratings for user_id. Each index represents the rating for i-th item. If the user was not present in the training set, a np.NaN vector is returned.

Parameters:user_id (int) Returns:ratings Return type:np.array containing all estimated items ratings for user_id.

save_model(filepath, overwrite=True, save_format='json')

Saves a model to a file. The model is synchronized before saving and can be reinstantiated in the exact same state, without any of the code used for model definition or fitting. :Parameters: * filepath (str) – Path where to save the model

• overwrite (bool, optional (default=True)) – Whether any existing model at the target location should be overwritten.
• save_format (str, optional (default=’json)) – Format used to save the models.

Examples

>>> from dislib.recommendation import ALS
>>> import numpy as np
>>> import dislib as ds
>>>  data = np.array([[0, 0, 5], [3, 0, 5], [3, 1, 2]])
>>> ratings = csr_matrix(data)
>>> train = ds.array(x=ratings, block_size=(1, 1))
>>> als = ALS(tol=0.01, random_state=666, n_f=5, verbose=False)
>>> als.fit(train)
>>> als.save_model("model_als")