def test_movie_lens_fit(self):
"""
This test checks whether the movielens getter works and that the resulting data is viable for fitting/testing a
TensorRec model.
"""
train_interactions, test_interactions, user_features, item_features = get_movielens_100k(
)
model = TensorRec()
model.fit(interactions=train_interactions,
user_features=user_features,
item_features=item_features)
predictions = model.predict(user_features=user_features,
item_features=item_features)
self.assertIsNotNone(predictions)
from tensorrec import TensorRec
from tensorrec.eval import fit_and_eval
from tensorrec.representation_graphs import (
LinearRepresentationGraph, NormalizedLinearRepresentationGraph
)
from tensorrec.loss_graphs import BalancedWMRBLossGraph
from test.datasets import get_movielens_100k
import logging
logging.getLogger().setLevel(logging.INFO)
# Load the movielens dataset
train_interactions, test_interactions, user_features, item_features, _ = get_movielens_100k(negative_value=0)
# Construct parameters for fitting
epochs = 500
alpha = 0.00001
n_components = 10
verbose = True
learning_rate = .01
n_sampled_items = int(item_features.shape[0] * .1)
fit_kwargs = {'epochs': epochs, 'alpha': alpha, 'verbose': verbose, 'learning_rate': learning_rate,
'n_sampled_items': n_sampled_items}
# Build two models -- one without an attention graph, one with a linear attention graph
model_without_attention = TensorRec(
n_components=10,
n_tastes=3,
user_repr_graph=NormalizedLinearRepresentationGraph(),
attention_graph=None,
import numpy as np
from tensorrec import TensorRec
from tensorrec.eval import precision_at_k, recall_at_k
from tensorrec.loss_graphs import BalancedWMRBLossGraph
from tensorrec.prediction_graphs import DotProductPredictionGraph
from tensorrec.representation_graphs import NormalizedLinearRepresentationGraph
from test.datasets import get_movielens_100k
import logging
logging.getLogger().setLevel(logging.INFO)
# Load the movielens dataset
train_interactions, test_interactions, user_features, item_features, item_titles = \
get_movielens_100k(negative_value=-1.0)
# Assemble parameters for fitting. 'epochs' is 1 in the fit_kwargs because we will be calling fit_partial 1000 times to
# run 1000 epochs.
epochs = 1000
fit_kwargs = {'epochs': 1, 'alpha': 0.0001, 'verbose': True, 'learning_rate': .01,
'n_sampled_items': int(item_features.shape[0] * .1)}
# Build the TensorRec model
model = TensorRec(n_components=2,
biased=True,
loss_graph=BalancedWMRBLossGraph(),
prediction_graph=DotProductPredictionGraph(),
user_repr_graph=NormalizedLinearRepresentationGraph(),
normalize_users=True,
normalize_items=True,
from tensorrec.representation_graphs import (
LinearRepresentationGraph, ReLURepresentationGraph,
NormalizedLinearRepresentationGraph)
from tensorrec.loss_graphs import WMRBLossGraph, BalancedWMRBLossGraph
from tensorrec.prediction_graphs import (DotProductPredictionGraph,
CosineSimilarityPredictionGraph,
EuclidianSimilarityPredictionGraph)
from tensorrec.util import append_to_string_at_point
from test.datasets import get_movielens_100k
import logging
logging.getLogger().setLevel(logging.INFO)
# Load the movielens dataset
train_interactions, test_interactions, user_features, item_features, _ = get_movielens_100k(
negative_value=0)
# Construct parameters for fitting
epochs = 300
alpha = 0.00001
n_components = 10
verbose = True
learning_rate = .01
n_sampled_items = int(item_features.shape[0] * .1)
biased = False
fit_kwargs = {
'epochs': epochs,
'alpha': alpha,
'verbose': verbose,
'learning_rate': learning_rate,
'n_sampled_items': n_sampled_items
def setUpClass(cls):
cls.movielens_100k = get_movielens_100k()
def setUpClass(cls):
cls.movielens_100k = get_movielens_100k()
import numpy as np
from tensorrec import TensorRec
from tensorrec.eval import precision_at_k, recall_at_k
from tensorrec.input_utils import create_tensorrec_dataset_from_sparse_matrix
from tensorrec.loss_graphs import BalancedWMRBLossGraph
from tensorrec.representation_graphs import ReLURepresentationGraph
from test.datasets import get_movielens_100k
import logging
logging.getLogger().setLevel(logging.INFO)
# Load the movielens dataset
train_interactions, test_interactions, user_features, item_features, item_titles = \
get_movielens_100k(negative_value=-1.0)
# Assemble parameters for fitting. 'epochs' is 1 in the fit_kwargs because we will be calling fit_partial 1000 times to
# run 1000 epochs.
epochs = 1000
fit_kwargs = {'epochs': 1, 'alpha': 0.0001, 'verbose': True, 'learning_rate': .01,
'n_sampled_items': int(item_features.shape[0] * .1)}
# Build the TensorRec model
model = TensorRec(n_components=2,
biased=False,
loss_graph=BalancedWMRBLossGraph(),
item_repr_graph=ReLURepresentationGraph(),
n_tastes=3)
# Make some random selections of movies and users we want to plot
import keras as ks
from tensorrec import TensorRec
from tensorrec.eval import fit_and_eval
from tensorrec.representation_graphs import AbstractKerasRepresentationGraph
from tensorrec.loss_graphs import SeparationDenseLossGraph
from test.datasets import get_movielens_100k
import logging
logging.getLogger().setLevel(logging.INFO)
train_interactions, test_interactions, user_features, item_features, _ = get_movielens_100k(
)
class ExampleKerasRepresentationGraph(AbstractKerasRepresentationGraph):
def create_layers(self, n_features, n_components):
return [
ks.layers.Dense(int(n_features / 2), activation='relu'),
ks.layers.Dense(n_components * 2, activation='relu'),
ks.layers.Dense(n_components, activation='tanh'),
]
model = TensorRec(n_components=10,
item_repr_graph=ExampleKerasRepresentationGraph(),
loss_graph=SeparationDenseLossGraph())
fit_kwargs = {'epochs': 1000, 'learning_rate': .001, 'verbose': True}