bstds commited on
Commit
6774c1a
1 Parent(s): 67deb2b

Create movielens.py

Browse files
Files changed (1) hide show
  1. movielens.py +90 -0
movielens.py ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/usr/bin/env python3
2
+ # -*- coding: utf-8 -*-
3
+ import os
4
+ from functools import partial
5
+ from pathlib import Path
6
+ from typing import Dict, Iterable
7
+
8
+ import datasets
9
+ from datasets import DatasetDict, DownloadManager, load_dataset
10
+ import pandas as pd
11
+
12
+
13
+ AVAILABLE_DATASETS = {
14
+ 'small': 'https://files.grouplens.org/datasets/movielens/ml-latest-small.zip',
15
+ 'full': 'https://files.grouplens.org/datasets/movielens/ml-latest.zip',
16
+ }
17
+
18
+
19
+ VERSION = datasets.Version("0.0.1")
20
+
21
+
22
+ class MovielensDataset(datasets.GeneratorBasedBuilder):
23
+ """MovielensDataset dataset."""
24
+
25
+ BUILDER_CONFIGS = [
26
+ datasets.BuilderConfig(
27
+ name=data_name, version=VERSION, description=f"{data_name} movielens dataset"
28
+ )
29
+ for data_name in AVAILABLE_DATASETS
30
+ ]
31
+
32
+ def _info(self) -> datasets.DatasetInfo:
33
+ return datasets.DatasetInfo(
34
+ description="",
35
+ features=datasets.Features(
36
+ {
37
+ "movieId": datasets.Value("string"),
38
+ "title": datasets.Value("string"),
39
+ "genres": datasets.Sequence(datasets.Value("string")),
40
+ "tag": datasets.Sequence(datasets.Value("string")),
41
+ }
42
+ ),
43
+ supervised_keys=None,
44
+ homepage="https://grouplens.org/datasets/movielens/latest/",
45
+ citation="",
46
+ )
47
+
48
+ def _split_generators(
49
+ self, dl_manager: DownloadManager
50
+ ) -> Iterable[datasets.SplitGenerator]:
51
+ downloader = partial(
52
+ lambda split: dl_manager.download_and_extract(
53
+ AVAILABLE_DATASETS[self.config.name]
54
+ )
55
+ )
56
+
57
+ folder = os.path.splitext(os.path.basename(AVAILABLE_DATASETS[self.config.name]))[
58
+ 0
59
+ ]
60
+
61
+ # There is no predefined train/val/test split for this dataset.
62
+ return [
63
+ datasets.SplitGenerator(
64
+ name=datasets.Split.TRAIN,
65
+ gen_kwargs={
66
+ "root_path": downloader("train"),
67
+ "split": "train",
68
+ "folder": folder,
69
+ },
70
+ ),
71
+ ]
72
+
73
+ def _generate_examples(
74
+ self, root_path: str, split: str, folder: str
75
+ ) -> Iterable[Dict]:
76
+ split_path = Path(root_path) / folder
77
+ movies_file = split_path / "movies.csv"
78
+ movies = pd.read_csv(movies_file, sep=',', encoding='utf-8')
79
+ movies['genres'] = movies['genres'].str.split('|')
80
+ tags_file = split_path / "tags.csv"
81
+ tags = pd.read_csv(tags_file, sep=',', encoding='utf-8')
82
+ tags = tags.groupby('movieId').agg({'tag': list}).reset_index()
83
+ movies = movies.merge(tags, on='movieId')
84
+ for idx, row in movies.iterrows():
85
+ yield idx, {
86
+ 'movieId': row['movieId'],
87
+ 'title': row['title'],
88
+ 'genres': row['genres'],
89
+ 'tag': row['tag'],
90
+ }