Update
Browse files- test-dataset-debug.py +63 -80
test-dataset-debug.py
CHANGED
@@ -1,45 +1,43 @@
|
|
1 |
-
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
from __future__ import absolute_import, division, print_function
|
4 |
|
5 |
-
import json
|
6 |
import os
|
7 |
import datetime
|
8 |
import pandas as pd
|
9 |
import numpy as np
|
10 |
from pathlib import Path
|
|
|
|
|
|
|
|
|
11 |
|
12 |
import datasets
|
13 |
|
14 |
|
15 |
-
# TODO: Add BibTeX citation
|
16 |
_CITATION = """\
|
17 |
-
@InProceedings{
|
18 |
-
title = {
|
19 |
-
authors={
|
20 |
-
}
|
21 |
-
year={2020}
|
22 |
}
|
23 |
"""
|
24 |
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
# # URLs for debugging
|
35 |
-
# _METADATA_URL = _DEBUG_METADATA_URL = "https://patentdiag.blob.core.windows.net/patent-data/metadata_debug-2021-02-10.feather"
|
36 |
-
# _DATA_URL = _DEBUG_DATA_URL = "https://patentdiag.blob.core.windows.net/patent-data/distilled_debug-2021-01-07.tar"
|
37 |
-
# _DATA_SUBFOLDER_NAME = _DATA_SUBFOLDER_NAME = 'debug_distilled'
|
38 |
-
|
39 |
-
# URLs for figuring out the Huggingface Hub
|
40 |
-
_METADATA_URL = "https://huggingface.co/datasets/greeneggsandyaml/test-dataset-debug/resolve/main/metadata--Jan2016--2021-02-10.feather"
|
41 |
-
_DATA_URL = "https://huggingface.co/datasets/greeneggsandyaml/test-dataset-debug/resolve/main/json-files-Jan2016.tar"
|
42 |
-
_DATA_SUBFOLDER_NAME = 'json-files-Jan2016'
|
43 |
|
44 |
RANDOM_STATE = 1729
|
45 |
|
@@ -73,8 +71,11 @@ class PatentsConfig(datasets.BuilderConfig):
|
|
73 |
|
74 |
def __init__(
|
75 |
self,
|
76 |
-
|
77 |
-
|
|
|
|
|
|
|
78 |
train_filing_start_date: str = None,
|
79 |
train_filing_end_date: str = None,
|
80 |
val_filing_start_date: str = None,
|
@@ -82,7 +83,6 @@ class PatentsConfig(datasets.BuilderConfig):
|
|
82 |
query_string: str = None,
|
83 |
val_set_balancer=False,
|
84 |
uniform_split=False,
|
85 |
-
train_only=False,
|
86 |
**kwargs
|
87 |
):
|
88 |
"""
|
@@ -92,6 +92,9 @@ class PatentsConfig(datasets.BuilderConfig):
|
|
92 |
train_filing_end_date.
|
93 |
|
94 |
Args:
|
|
|
|
|
|
|
95 |
ipcr_label: International Patent Classification code
|
96 |
cpc_label: Cooperative Patent Classification code
|
97 |
train_filing_start_date: Start date for patents in train set (and val set if random split is used)
|
@@ -101,6 +104,9 @@ class PatentsConfig(datasets.BuilderConfig):
|
|
101 |
**kwargs: keyword arguments forwarded to super.
|
102 |
"""
|
103 |
super().__init__(**kwargs)
|
|
|
|
|
|
|
104 |
self.ipcr_label = ipcr_label
|
105 |
self.cpc_label = cpc_label
|
106 |
self.train_filing_start_date = train_filing_start_date
|
@@ -110,11 +116,10 @@ class PatentsConfig(datasets.BuilderConfig):
|
|
110 |
self.query_string = query_string
|
111 |
self.val_set_balancer = val_set_balancer
|
112 |
self.uniform_split = uniform_split
|
113 |
-
self.train_only = train_only
|
114 |
|
115 |
|
116 |
class Patents(datasets.GeneratorBasedBuilder):
|
117 |
-
|
118 |
|
119 |
VERSION = datasets.Version("1.0.1")
|
120 |
|
@@ -122,10 +127,22 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
122 |
# If you don't want/need to define several sub-sets in your dataset,
|
123 |
# just remove the BUILDER_CONFIG_CLASS and the BUILDER_CONFIGS attributes.
|
124 |
BUILDER_CONFIG_CLASS = PatentsConfig
|
125 |
-
|
126 |
-
|
127 |
-
|
128 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
129 |
|
130 |
def _info(self):
|
131 |
return datasets.DatasetInfo(
|
@@ -139,8 +156,7 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
139 |
# specify them here. They'll be used if as_supervised=True in
|
140 |
# builder.as_dataset.
|
141 |
supervised_keys=("claims", "decision"),
|
142 |
-
|
143 |
-
homepage="https://huggingface.co/great-new-dataset",
|
144 |
citation=_CITATION,
|
145 |
)
|
146 |
|
@@ -149,24 +165,16 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
149 |
print(f'Loading dataset with config: {self.config}')
|
150 |
|
151 |
# Download metadata
|
152 |
-
# NOTE:
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
else:
|
157 |
-
print(f'Using metadata file: {self.config.data_files}')
|
158 |
-
metadata_file = Path(self.config.data_files)
|
159 |
|
160 |
# Download data
|
161 |
-
# NOTE:
|
162 |
-
# json file per patent application
|
163 |
-
|
164 |
-
|
165 |
-
json_dir = Path(dl_manager.download_and_extract(_DATA_URL))
|
166 |
-
# NOTE: The extracted path contains a subfolder
|
167 |
-
json_dir = json_dir / _DATA_SUBFOLDER_NAME
|
168 |
-
else:
|
169 |
-
json_dir = Path(self.config.data_dir)
|
170 |
|
171 |
# Load metadata file
|
172 |
print(f'Reading metadata file: {metadata_file}')
|
@@ -181,32 +189,9 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
181 |
df = df[df['main_cpc_label'].str.startswith(self.config.cpc_label)]
|
182 |
|
183 |
# Filter metadata based on arbitrary query string
|
184 |
-
# TODO(suproteem): remove for production
|
185 |
if self.config.query_string:
|
186 |
df = df.query(self.config.query_string)
|
187 |
|
188 |
-
|
189 |
-
|
190 |
-
# Return only one dataset
|
191 |
-
if self.config.train_only:
|
192 |
-
if self.config.train_filing_start_date:
|
193 |
-
print(f'Filtering by train filing start date: {self.config.train_filing_start_date}')
|
194 |
-
df = df[df['filing_date'] >= self.config.train_filing_start_date]
|
195 |
-
if self.config.train_filing_end_date:
|
196 |
-
print(f'Filtering by train filing end date: {self.config.train_filing_end_date}')
|
197 |
-
df = df[df['filing_date'] <= self.config.train_filing_end_date]
|
198 |
-
|
199 |
-
return [
|
200 |
-
datasets.SplitGenerator(
|
201 |
-
name=datasets.Split.TRAIN,
|
202 |
-
gen_kwargs=dict( # kwargs passed to _generate_examples
|
203 |
-
df=df,
|
204 |
-
json_dir=json_dir,
|
205 |
-
split='train',
|
206 |
-
),
|
207 |
-
)
|
208 |
-
]
|
209 |
-
|
210 |
# Train-validation split (either uniform or by date)
|
211 |
if self.config.uniform_split:
|
212 |
|
@@ -242,7 +227,7 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
242 |
(df['filing_date'] < self.config.val_filing_end_date)
|
243 |
]
|
244 |
|
245 |
-
# TODO:
|
246 |
if self.config.val_set_balancer:
|
247 |
rejected_df = val_df[val_df.status == 'REJECTED']
|
248 |
num_rejected = len(rejected_df)
|
@@ -259,7 +244,7 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
259 |
return [
|
260 |
datasets.SplitGenerator(
|
261 |
name=datasets.Split.TRAIN,
|
262 |
-
gen_kwargs=dict( # kwargs passed to _generate_examples
|
263 |
df=train_df,
|
264 |
json_dir=json_dir,
|
265 |
split='train',
|
@@ -311,6 +296,4 @@ class Patents(datasets.GeneratorBasedBuilder):
|
|
311 |
'date_published': patent['date_published'],
|
312 |
'examiner_id': patent['examiner_id'],
|
313 |
"ipc_label": patent["main_ipcr_label"],
|
314 |
-
# "all_cpc_labels": patent["cpc_labels"], # these are lists, ignoring for now
|
315 |
-
# 'inventor_list': patent['inventor_list'], # these are lists, ignoring for now
|
316 |
}
|
|
|
1 |
+
"""
|
2 |
+
The Harvard USPTO Patent Dataset (HUPD) is a large-scale, well-structured, and multi-purpose corpus
|
3 |
+
of English-language patent applications filed to the United States Patent and Trademark Office (USPTO)
|
4 |
+
between 2004 and 2018. With more than 4.5 million patent documents, HUPD is two to three times larger
|
5 |
+
than comparable corpora. Unlike other NLP patent datasets, HUPD contains the inventor-submitted versions
|
6 |
+
of patent applications, not the final versions of granted patents, allowing us to study patentability at
|
7 |
+
the time of filing using NLP methods for the first time.
|
8 |
+
"""
|
9 |
|
10 |
from __future__ import absolute_import, division, print_function
|
11 |
|
|
|
12 |
import os
|
13 |
import datetime
|
14 |
import pandas as pd
|
15 |
import numpy as np
|
16 |
from pathlib import Path
|
17 |
+
try:
|
18 |
+
import ujson as json
|
19 |
+
except:
|
20 |
+
import json
|
21 |
|
22 |
import datasets
|
23 |
|
24 |
|
|
|
25 |
_CITATION = """\
|
26 |
+
@InProceedings{suzgun2021:hupd,
|
27 |
+
title = {The Harvard USPTO Patent Dataset},
|
28 |
+
authors={Mirac Suzgun and Suproteem Sarkar and Luke Melas-Kyriazi and Scott Kominers and Stuart Shieber},
|
29 |
+
year={2021}
|
|
|
30 |
}
|
31 |
"""
|
32 |
|
33 |
+
_DESCRIPTION = """
|
34 |
+
The Harvard USPTO Patent Dataset (HUPD) is a large-scale, well-structured, and multi-purpose corpus
|
35 |
+
of English-language patent applications filed to the United States Patent and Trademark Office (USPTO)
|
36 |
+
between 2004 and 2018. With more than 4.5 million patent documents, HUPD is two to three times larger
|
37 |
+
than comparable corpora. Unlike other NLP patent datasets, HUPD contains the inventor-submitted versions
|
38 |
+
of patent applications, not the final versions of granted patents, allowing us to study patentability at
|
39 |
+
the time of filing using NLP methods for the first time.
|
40 |
+
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
|
42 |
RANDOM_STATE = 1729
|
43 |
|
|
|
71 |
|
72 |
def __init__(
|
73 |
self,
|
74 |
+
metadata_url: str,
|
75 |
+
data_url: str,
|
76 |
+
data_dir: str,
|
77 |
+
ipcr_label: str = None,
|
78 |
+
cpc_label: str = None,
|
79 |
train_filing_start_date: str = None,
|
80 |
train_filing_end_date: str = None,
|
81 |
val_filing_start_date: str = None,
|
|
|
83 |
query_string: str = None,
|
84 |
val_set_balancer=False,
|
85 |
uniform_split=False,
|
|
|
86 |
**kwargs
|
87 |
):
|
88 |
"""
|
|
|
92 |
train_filing_end_date.
|
93 |
|
94 |
Args:
|
95 |
+
metadata_url: `string`, url from which to download the metadata file
|
96 |
+
data_url: `string`, url from which to download the json files
|
97 |
+
data_dir: `string`, folder (in cache) in which downloaded json files are stored
|
98 |
ipcr_label: International Patent Classification code
|
99 |
cpc_label: Cooperative Patent Classification code
|
100 |
train_filing_start_date: Start date for patents in train set (and val set if random split is used)
|
|
|
104 |
**kwargs: keyword arguments forwarded to super.
|
105 |
"""
|
106 |
super().__init__(**kwargs)
|
107 |
+
self.metadata_url = metadata_url
|
108 |
+
self.data_url = data_url
|
109 |
+
self.data_dir = data_dir
|
110 |
self.ipcr_label = ipcr_label
|
111 |
self.cpc_label = cpc_label
|
112 |
self.train_filing_start_date = train_filing_start_date
|
|
|
116 |
self.query_string = query_string
|
117 |
self.val_set_balancer = val_set_balancer
|
118 |
self.uniform_split = uniform_split
|
|
|
119 |
|
120 |
|
121 |
class Patents(datasets.GeneratorBasedBuilder):
|
122 |
+
_DESCRIPTION
|
123 |
|
124 |
VERSION = datasets.Version("1.0.1")
|
125 |
|
|
|
127 |
# If you don't want/need to define several sub-sets in your dataset,
|
128 |
# just remove the BUILDER_CONFIG_CLASS and the BUILDER_CONFIGS attributes.
|
129 |
BUILDER_CONFIG_CLASS = PatentsConfig
|
130 |
+
BUILDER_CONFIGS = [
|
131 |
+
PatentsConfig(
|
132 |
+
name="sample",
|
133 |
+
description="Patent data from January 2016, for debugging",
|
134 |
+
metadata_url="https://huggingface.co/datasets/greeneggsandyaml/test-dataset-debug/resolve/main/metadata--Jan2016--2021-02-10.feather",
|
135 |
+
data_url="https://huggingface.co/datasets/greeneggsandyaml/test-dataset-debug/resolve/main/json-files-Jan2016.tar",
|
136 |
+
data_dir="json-files-Jan2016",
|
137 |
+
),
|
138 |
+
PatentsConfig(
|
139 |
+
name="all",
|
140 |
+
description="Patent data from January 2016, for debugging",
|
141 |
+
metadata_url="https://patentdiag.blob.core.windows.net/patent-data/metadata-2021-02-10.feather",
|
142 |
+
data_url="https://patentdiag.blob.core.windows.net/patent-data/distilled-2021-01-07.tar",
|
143 |
+
data_dir="distilled",
|
144 |
+
),
|
145 |
+
]
|
146 |
|
147 |
def _info(self):
|
148 |
return datasets.DatasetInfo(
|
|
|
156 |
# specify them here. They'll be used if as_supervised=True in
|
157 |
# builder.as_dataset.
|
158 |
supervised_keys=("claims", "decision"),
|
159 |
+
homepage="https://github.com/suzgunmirac/hupd",
|
|
|
160 |
citation=_CITATION,
|
161 |
)
|
162 |
|
|
|
165 |
print(f'Loading dataset with config: {self.config}')
|
166 |
|
167 |
# Download metadata
|
168 |
+
# NOTE: Metadata is stored as a Pandas DataFrame in Apache Feather format
|
169 |
+
metadata_file = dl_manager.download_and_extract(self.config.metadata_url)
|
170 |
+
metadata_file = Path(self.config.data_files)
|
171 |
+
print(f'Using metadata file: {self.config.data_files}')
|
|
|
|
|
|
|
172 |
|
173 |
# Download data
|
174 |
+
# NOTE: The extracted path contains a subfolder, data_dir. This directory holds
|
175 |
+
# a large number of json files (one json file per patent application).
|
176 |
+
download_dir = dl_manager.download_and_extract(self.config.data_url)
|
177 |
+
json_dir = os.path.join(download_dir, self.config.data_dir)
|
|
|
|
|
|
|
|
|
|
|
178 |
|
179 |
# Load metadata file
|
180 |
print(f'Reading metadata file: {metadata_file}')
|
|
|
189 |
df = df[df['main_cpc_label'].str.startswith(self.config.cpc_label)]
|
190 |
|
191 |
# Filter metadata based on arbitrary query string
|
|
|
192 |
if self.config.query_string:
|
193 |
df = df.query(self.config.query_string)
|
194 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
195 |
# Train-validation split (either uniform or by date)
|
196 |
if self.config.uniform_split:
|
197 |
|
|
|
227 |
(df['filing_date'] < self.config.val_filing_end_date)
|
228 |
]
|
229 |
|
230 |
+
# TODO: We can probably make this step faster
|
231 |
if self.config.val_set_balancer:
|
232 |
rejected_df = val_df[val_df.status == 'REJECTED']
|
233 |
num_rejected = len(rejected_df)
|
|
|
244 |
return [
|
245 |
datasets.SplitGenerator(
|
246 |
name=datasets.Split.TRAIN,
|
247 |
+
gen_kwargs=dict( # these kwargs are passed to _generate_examples
|
248 |
df=train_df,
|
249 |
json_dir=json_dir,
|
250 |
split='train',
|
|
|
296 |
'date_published': patent['date_published'],
|
297 |
'examiner_id': patent['examiner_id'],
|
298 |
"ipc_label": patent["main_ipcr_label"],
|
|
|
|
|
299 |
}
|