# Copyright 2023-2024 MTS (Mobile Telesystems)
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""RandomSplitter."""
import typing as tp
from typing import Optional
import numpy as np
import pandas as pd
from rectools.dataset import Interactions
from rectools.model_selection.splitter import Splitter
[docs]class RandomSplitter(Splitter):
"""
Slitter for cross-validation by random.
Generate train and test folds with fixed test part ratio
without intersections between test folds.
Random splitting is applied to interactions.
Users and items are not taken into account while preparing splits.
It is also possible to exclude cold users and items
and already seen items.
Parameters
----------
test_fold_frac : float
Relative size of test part, must be between 0. and 1.
n_splits : int, default 1
Number of test folds.
random_state : int, default None,
Controls randomness of each fold. Pass an int to get reproducible result across multiple `split` calls.
filter_cold_users : bool, default ``True``
If `True`, users that are not present in train will be excluded from test.
WARNING: both cold and warm users will be excluded from test.
filter_cold_items : bool, default ``True``
If `True`, items that are not present in train will be excluded from test.
WARNING: both cold and warm items will be excluded from test.
filter_already_seen : bool, default ``True``
If `True`, pairs (user, item) that are present in train will be excluded from test.
Examples
--------
>>> from rectools import Columns
>>> df = pd.DataFrame(
... [
... [1, 2, 1, "2021-09-01"], # 0
... [2, 1, 1, "2021-09-02"], # 1
... [2, 3, 1, "2021-09-03"], # 2
... [3, 2, 1, "2021-09-03"], # 3
... [3, 3, 1, "2021-09-04"], # 4
... [3, 4, 1, "2021-09-04"], # 5
... [1, 2, 1, "2021-09-05"], # 6
... [4, 2, 1, "2021-09-05"], # 7
... ],
... columns=[Columns.User, Columns.Item, Columns.Weight, Columns.Datetime],
... ).astype({Columns.Datetime: "datetime64[ns]"})
>>> interactions = Interactions(df)
>>>
>>> splitter = RandomSplitter(test_fold_frac=0.25, random_state=42, n_splits=2, filter_cold_users=False,
... filter_cold_items=False, filter_already_seen=False)
>>> for train_ids, test_ids, _ in splitter.split(interactions):
... print(train_ids, test_ids)
[2 7 6 1 5 0] [3 4]
[3 4 6 1 5 0] [2 7]
>>>
>>> splitter = RandomSplitter(test_fold_frac=0.25, random_state=42, n_splits=2, filter_cold_users=True,
... filter_cold_items=True, filter_already_seen=True)
>>> for train_ids, test_ids, _ in splitter.split(interactions):
... print(train_ids, test_ids)
[2 7 6 1 5 0] [3 4]
[3 4 6 1 5 0] [2]
"""
def __init__(
self,
test_fold_frac: float,
n_splits: int = 1,
random_state: Optional[int] = None,
filter_cold_users: bool = True,
filter_cold_items: bool = True,
filter_already_seen: bool = True,
) -> None:
if test_fold_frac <= 0.0 or test_fold_frac >= 1.0:
raise ValueError("Value of test_fold_frac must be between 0 and 1")
if test_fold_frac * n_splits > 1:
raise ValueError(f"Impossible to create {n_splits} non-overlapping folds {test_fold_frac:.1%} each")
super().__init__(filter_cold_users, filter_cold_items, filter_already_seen)
self.random_state = random_state
self.n_splits = n_splits
self.test_fold_frac = test_fold_frac
def _split_without_filter(
self,
interactions: Interactions,
collect_fold_stats: bool = False,
) -> tp.Iterator[tp.Tuple[np.ndarray, np.ndarray, tp.Dict[str, tp.Any]]]:
rng = np.random.default_rng(self.random_state)
df = interactions.df
idx = pd.RangeIndex(0, len(df))
test_fold_size = int(round(self.test_fold_frac * len(df)))
if test_fold_size == 0:
raise ValueError(
f"Length of interactions ({len(df)}) with test_fold_frac={self.test_fold_frac} leads to empty test part"
)
if test_fold_size == len(df):
raise ValueError(
f"Length of interactions ({len(df)}) with test_fold_frac={self.test_fold_frac} "
"leads to empty train part: all interactions are related to the test"
)
if self.n_splits * test_fold_size > len(df):
raise ValueError(
f"Impossible to create {self.n_splits} non-overlapping folds "
f"with size {test_fold_size} from {len(df)} interactions"
)
shuffled_idx = rng.permutation(idx)
for i_split in range(self.n_splits):
left = i_split * test_fold_size
right = (i_split + 1) * test_fold_size
test_idx = shuffled_idx[left:right]
train_idx = np.concatenate((shuffled_idx[:left], shuffled_idx[right:]))
split_info = {"i_split": i_split}
yield train_idx, test_idx, split_info