def load_interaction(self):
"""Load the user-item interaction.
Load the interaction from the processed file(Need to preprocess the raw file before loading)
"""
processed_file_path = os.path.join(
self.processed_path, f"{self.dataset_name}_interaction.npz")
if not os.path.exists(os.path.join(processed_file_path)):
try:
self.preprocess()
except FileNotFoundError:
print("origin file is broken, re-download it")
raw_file_path = os.path.join(self.raw_path,
f"{self.dataset_name}.zip")
os.remove(raw_file_path)
self.download()
finally:
self.preprocess()
data = get_dataframe_from_npz(processed_file_path)
print("-" * 80)
print("Loaded raw interaction statistics")
print(
tabulate(
data.agg(["count", "nunique"]),
headers=data.columns,
tablefmt="psql",
disable_numparse=True,
))
print("-" * 80)
return data
def load_interaction(self):
"""Load the user-item interaction
Load the interaction from the processed file(Need to preprocess the raw file before loading)
"""
processed_file_path = os.path.join(
self.processed_path, f"{self.dataset_name}_interaction.npz")
if not os.path.exists(os.path.join(processed_file_path)):
self.preprocess()
data = get_dataframe_from_npz(processed_file_path)
print("-" * 80)
print("raw interaction statistics")
print(
tabulate(
data.agg(["count", "nunique"]),
headers=data.columns,
tablefmt="psql",
disable_numparse=True,
))
print("-" * 80)
return data
def load_split_data(path, n_test=10):
"""Load split DataFrame from a specified path.
Args:
path (string): split data path.
n_test: number of testing and validation datasets.
If n_test==0, will load the original (no negative items) valid and test datasets.
Returns:
(DataFrame, list(DataFrame), list(DataFrame)): DataFrame of training interaction,
DataFrame list of validation interaction,
DataFrame list of testing interaction,
"""
train_file = os.path.join(path, "train.npz")
train_data = get_dataframe_from_npz(train_file)
print("-" * 80)
print("Loaded training set statistics")
print(
tabulate(
train_data.agg(["count", "nunique"]),
headers=train_data.columns,
tablefmt="psql",
disable_numparse=True,
)
)
if not n_test:
valid_df = get_dataframe_from_npz(os.path.join(path, "valid.npz"))
test_df = get_dataframe_from_npz(os.path.join(path, "test.npz"))
print("Loaded validation set statistics")
print(
tabulate(
valid_df.agg(["count", "nunique"]),
headers=valid_df.columns,
tablefmt="psql",
disable_numparse=True,
)
)
print("Loaded testing set statistics")
print(
tabulate(
test_df.agg(["count", "nunique"]),
headers=test_df.columns,
tablefmt="psql",
disable_numparse=True,
)
)
print("-" * 80)
return train_data, valid_df, test_df
valid_data_li = []
test_data_li = []
for i in range(n_test):
valid_df = get_dataframe_from_npz(os.path.join(path, f"valid_{i}.npz"))
valid_data_li.append(valid_df)
if i == 0:
print(f"valid_data_{i} statistics")
print(
tabulate(
valid_df.agg(["count", "nunique"]),
headers=valid_df.columns,
tablefmt="psql",
disable_numparse=True,
)
)
test_df = get_dataframe_from_npz(os.path.join(path, f"test_{i}.npz"))
test_data_li.append(test_df)
if i == 0:
print(f"test_data_{i} statistics")
print(
tabulate(
test_df.agg(["count", "nunique"]),
headers=test_df.columns,
tablefmt="psql",
disable_numparse=True,
)
)
print("-" * 80)
return train_data, valid_data_li, test_data_li