Modules

Abstractions

class reclist.abstractions.RecDataset(force_download=False)

Bases: ABC

Implements an abstract class for the dataset

property catalog

abstract load()

Abstract method that should implement dataset loading :return:

property x_test

property x_train

property y_test

property y_train

class reclist.abstractions.RecList(model: RecModel, dataset: RecDataset, y_preds: Optional[list] = None)

Bases: ABC

META_DATA_FOLDER = '.reclist'

dump_results_to_json(test_results: list, report_path: str, epoch_time_ms: int)

generate_report(epoch_time_ms: int)

get_tests()

Helper to extract methods decorated with rec_test

property rec_tests

store_artifacts(report_path: str)

property test_data

property test_results

train_dense_repr(type_name: str, type_fn)

Train a dense representation over a type of meta-data & store into object

class reclist.abstractions.RecModel(model=None)

Bases: ABC

Abstract class for recommendation model

property model

abstract predict(prediction_input: list, *args, **kwargs)

The predict function should implement the behaviour of the model at inference time.

Parameters

• prediction_input – the input that is used to to do the prediction

• args –

• kwargs –

Returns

reclist.abstractions.rec_test(test_type: str)

Rec test decorator

Datasets

Data sets currently available

class reclist.datasets.CoveoDataset(**kwargs)

Bases: RecDataset

Coveo SIGIR data challenge dataset

load()

Abstract method that should implement dataset loading :return:

class reclist.datasets.MovieLensDataset(**kwargs)

Bases: RecDataset

MovieLens 25M Dataset

Reference: https://files.grouplens.org/datasets/movielens/ml-25m-README.html

load()

Abstract method that should implement dataset loading :return:

class reclist.datasets.SpotifyDataset(**kwargs)

Bases: RecDataset

load()

Abstract method that should implement dataset loading :return:

load_spotify_playlist_dataset()

Metrics

Metrics