def make_temporal_basket_split(self,
data=None,
test_rate=0.1,
n_negative=100,
by_user=False,
n_test=10):
"""Generate split data with temporal_basket_split.
Generate split data with temporal_basket_split method.
Args:
data (DataFrame): DataFrame to be split.
- Default is None. It will load the raw interaction, with a default filter
```
data = filter_user_item_order(data, min_u_c=10, min_o_c=10, min_i_c=10)
```
- Users can specify their filtered data by using filter methods in data_split.py
test_rate: percentage of the test data. Note that percentage of the validation data will be the same as testing.
n_negative: Number of negative samples for testing and validation data.
by_user: bool. Default False.
- True: user-based split,
- False: global split,
n_test: int. Default 10. The number of testing and validation copies.
Returns:
train_data (DataFrame): Interaction for training.
valid_data list(DataFrame): List of interactions for validation
test_data list(DataFrame): List of interactions for testing
"""
if data is None:
data = self.load_interaction()
data = filter_user_item_order(data,
min_u_c=10,
min_o_c=10,
min_i_c=10)
if not isinstance(data, pd.DataFrame):
raise RuntimeError("data is not a type of DataFrame")
if DEFAULT_TIMESTAMP_COL not in data.columns:
raise RuntimeError("This dataset doesn't have an TIMESTAMP_COL")
if DEFAULT_ORDER_COL not in data.columns:
raise RuntimeError("This dataset doesn't have an ORDER_COL")
result = split_data(
data,
split_type="temporal_basket",
test_rate=test_rate,
n_negative=n_negative,
save_dir=self.processed_path,
by_user=by_user,
n_test=n_test,
)
return result
def make_leave_one_basket(self,
data=None,
random=False,
n_negative=100,
n_test=10):
"""Generate split data with leave_one_basket.
Generate split data with leave_one_basket method.
Args:
data (DataFrame): DataFrame to be split.
- Default is None. It will load the raw interaction, with a default filter
'''
filter_user_item(data, min_u_c=0, min_o_c=3, min_i_c=0)
'''
- Users can specify their filtered data by using filter methods in data_split.py
random: bool. Whether randomly leave one basket as testing.
n_negative: Number of negative samples for testing and validation data.
n_test: int. Default 10. The number of testing and validation copies.
Returns:
train_data (DataFrame): Interaction for training.
valid_data list(DataFrame): List of interactions for validation
test_data list(DataFrame): List of interactions for testing
"""
if data is None:
data = self.load_interaction()
data = filter_user_item_order(data,
min_u_c=0,
min_o_c=3,
min_i_c=0)
if not isinstance(data, pd.DataFrame):
raise RuntimeError("data is not a type of DataFrame")
if DEFAULT_TIMESTAMP_COL not in data.columns:
raise RuntimeError("This dataset doesn't have an TIMESTAMP_COL")
if DEFAULT_ORDER_COL not in data.columns:
raise RuntimeError("This dataset doesn't have an ORDER_COL")
result = split_data(
data,
split_type="leave_one_basket",
test_rate=0,
random=random,
n_negative=n_negative,
save_dir=self.processed_path,
n_test=n_test,
)
return result
