# Copyright (c) Recommenders contributors.
# Licensed under the MIT License.
import numpy as np
import pandas as pd
from scipy.sparse import csr_matrix
from recommenders.utils.constants import (
DEFAULT_ITEM_COL,
DEFAULT_USER_COL,
DEFAULT_RATING_COL,
DEFAULT_TIMESTAMP_COL,
)
[docs]class RLRMCdataset(object):
"""RLRMC dataset implementation. Creates sparse data structures for RLRMC algorithm."""
[docs] def __init__(
self,
train,
validation=None,
test=None,
mean_center=True,
col_user=DEFAULT_USER_COL,
col_item=DEFAULT_ITEM_COL,
col_rating=DEFAULT_RATING_COL,
col_timestamp=DEFAULT_TIMESTAMP_COL,
# seed=42,
):
"""Initialize parameters.
Args:
train (pandas.DataFrame: training data with at least columns (col_user, col_item, col_rating)
validation (pandas.DataFrame): validation data with at least columns (col_user, col_item, col_rating). validation can be None, if so, we only process the training data
mean_center (bool): flag to mean center the ratings in train (and validation) data
col_user (str): user column name
col_item (str): item column name
col_rating (str): rating column name
col_timestamp (str): timestamp column name
"""
# initialize user and item index
self.user_idx = None
self.item_idx = None
# get col name of user, item and rating
self.col_user = col_user
self.col_item = col_item
self.col_rating = col_rating
self.col_timestamp = col_timestamp
# set random seed
# random.seed(seed)
# data preprocessing for training and validation data
self._data_processing(train, validation, test, mean_center)
def _data_processing(self, train, validation=None, test=None, mean_center=True):
"""Process the dataset to reindex userID and itemID
Args:
train (pandas.DataFrame): training data with at least columns (col_user, col_item, col_rating)
validation (pandas.DataFrame): validation data with at least columns (col_user, col_item, col_rating). validation can be None, if so, we only process the training data
mean_center (bool): flag to mean center the ratings in train (and validation) data
Returns:
list: train and validation pandas.DataFrame Dataset, which have been reindexed.
"""
# Data processing and reindexing code is adopted from https://github.com/Microsoft/Recommenders/blob/main/recommenders/models/ncf/dataset.py
# If validation dataset is None
df = train if validation is None else train.append(validation)
df = df if test is None else df.append(test)
# Reindex user and item index
if self.user_idx is None:
# Map user id
user_idx = df[[self.col_user]].drop_duplicates().reindex()
user_idx[self.col_user + "_idx"] = np.arange(len(user_idx))
self.n_users = len(user_idx)
self.user_idx = user_idx
self.user2id = dict(
zip(user_idx[self.col_user], user_idx[self.col_user + "_idx"])
)
self.id2user = {self.user2id[k]: k for k in self.user2id}
if self.item_idx is None:
# Map item id
item_idx = df[[self.col_item]].drop_duplicates()
item_idx[self.col_item + "_idx"] = np.arange(len(item_idx))
self.n_items = len(item_idx)
self.item_idx = item_idx
self.item2id = dict(
zip(item_idx[self.col_item], item_idx[self.col_item + "_idx"])
)
self.id2item = {self.item2id[k]: k for k in self.item2id}
df_train = self._reindex(train)
d = len(user_idx) # number of rows
T = len(item_idx) # number of columns
rows_train = df_train["userID"].values
cols_train = df_train["itemID"].values
entries_omega = df_train["rating"].values
if mean_center:
train_mean = np.mean(entries_omega)
else:
train_mean = 0.0
entries_train = entries_omega - train_mean
self.model_param = {"num_row": d, "num_col": T, "train_mean": train_mean}
self.train = csr_matrix(
(entries_train.T.ravel(), (rows_train, cols_train)), shape=(d, T)
)
if validation is not None:
df_validation = self._reindex(validation)
rows_validation = df_validation["userID"].values
cols_validation = df_validation["itemID"].values
entries_validation = df_validation["rating"].values - train_mean
self.validation = csr_matrix(
(entries_validation.T.ravel(), (rows_validation, cols_validation)),
shape=(d, T),
)
else:
self.validation = None
def _reindex(self, df):
"""Process dataset to reindex userID and itemID
Args:
df (pandas.DataFrame): dataframe with at least columns (col_user, col_item, col_rating)
Returns:
list: train and validation pandas.DataFrame Dataset, which have been reindexed.
"""
# If validation dataset is None
if df is None:
return None
# Map user_idx and item_idx
df = pd.merge(df, self.user_idx, on=self.col_user, how="left")
df = pd.merge(df, self.item_idx, on=self.col_item, how="left")
# Select relevant columns
df_reindex = df[
[self.col_user + "_idx", self.col_item + "_idx", self.col_rating]
]
df_reindex.columns = [self.col_user, self.col_item, self.col_rating]
return df_reindex