jmc255 commited on
Commit
a92ee6d
1 Parent(s): 93e0813

Upload aphantasia_drawing_dataset.py

Browse files
Files changed (1) hide show
  1. aphantasia_drawing_dataset.py +5 -47
aphantasia_drawing_dataset.py CHANGED
@@ -7,7 +7,7 @@ Original file is located at
7
  https://colab.research.google.com/drive/1DYVroeFqoNK7DDiw_3OIczPeqEME-rbh
8
  """
9
 
10
- # !pip install -q datasets
11
 
12
  # Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
13
  #
@@ -27,13 +27,9 @@ Original file is located at
27
 
28
 
29
 
30
- import base64
31
- from PIL import Image
32
  import pandas as pd
33
  import numpy as np
34
  import io
35
- import json
36
- import os
37
  from typing import List
38
  import datasets
39
  import logging
@@ -87,7 +83,7 @@ class AphantasiaDrawingDataset(datasets.GeneratorBasedBuilder):
87
  description=_DESCRIPTION,
88
  features=datasets.Features({
89
  "subject_id": datasets.Value("int32"),
90
- "treatment": datasets.Value("string"),#datasets.ClassLabel(names=["aphantasia", "control"]),
91
  "demographics": {
92
  "country": datasets.Value("string"),
93
  "age": datasets.Value("int32"),
@@ -145,49 +141,10 @@ class AphantasiaDrawingDataset(datasets.GeneratorBasedBuilder):
145
  image = Image.open(image_buffer)
146
  return image
147
  return None
148
-
149
  with open(filepath, "rb") as subjects_file:
150
  subjects_data = pd.read_parquet(subjects_file)
151
- #idx = 0
152
  for idx, sub_row in subjects_data.iterrows():
153
- # for room in subjects_data[sub]["drawings"].keys():
154
- # if subjects_data[sub]["drawings"][room]["perception"] != "":
155
- # img_byt = base64.b64decode(subjects_data[sub]["drawings"][room]["perception"])
156
- # img = Image.open(io.BytesIO(img_byt))
157
- # subjects_data[sub]["drawings"][room]["perception"] = img
158
- # else:
159
- # subjects_data[sub]["drawings"][room]["perception"] = None
160
-
161
- # if subjects_data[sub]["drawings"][room]["memory"] != "":
162
- # img_byt = base64.b64decode(subjects_data[sub]["drawings"][room]["memory"])
163
- # img = Image.open(io.BytesIO(img_byt))
164
- # subjects_data[sub]["drawings"][room]["memory"] = img
165
- # else:
166
- # subjects_data[sub]["drawings"][room]["memory"] = None
167
-
168
- # for room in subjects_data[sub]["image"].keys():
169
- # img_byt = base64.b64decode(subjects_data[sub]["image"][room])
170
- # img = Image.open(io.BytesIO(img_byt))
171
- # subjects_data[sub]["image"][room] = img.resize((500,500))
172
- # idx += 1
173
-
174
- # try:
175
- # age = int(sub_row["demographics.age"])
176
- # except:
177
- # age = -1
178
- # try:
179
- # art_ability = int(sub_row["demographics.art_ability"])
180
- # except:
181
- # art_ability = -1
182
- # try:
183
- # vviq_score = int(sub_row["demographics.vviq_score"])
184
- # except:
185
- # vviq_score = -1
186
- # try:
187
- # osiq_score = int(sub_row["demographics.osiq_score"])
188
- # except:
189
- # osiq_score = -1
190
- # age = int(sub_row["demographics.age"]) if sub_row["demographics.age"] and sub_row["demographics.age"] is not np.nan else None
191
  yield idx, {
192
  "subject_id": sub_row["subject_id"],
193
  "treatment": sub_row["treatment"],
@@ -224,4 +181,5 @@ class AphantasiaDrawingDataset(datasets.GeneratorBasedBuilder):
224
  "livingroom": byt_to_image(sub_row["image.livingroom"]),
225
  "bedroom": byt_to_image(sub_row["image.bedroom"])
226
  }
227
- }
 
 
7
  https://colab.research.google.com/drive/1DYVroeFqoNK7DDiw_3OIczPeqEME-rbh
8
  """
9
 
10
+
11
 
12
  # Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
13
  #
 
27
 
28
 
29
 
 
 
30
  import pandas as pd
31
  import numpy as np
32
  import io
 
 
33
  from typing import List
34
  import datasets
35
  import logging
 
83
  description=_DESCRIPTION,
84
  features=datasets.Features({
85
  "subject_id": datasets.Value("int32"),
86
+ "treatment": datasets.Value("string"),
87
  "demographics": {
88
  "country": datasets.Value("string"),
89
  "age": datasets.Value("int32"),
 
141
  image = Image.open(image_buffer)
142
  return image
143
  return None
144
+
145
  with open(filepath, "rb") as subjects_file:
146
  subjects_data = pd.read_parquet(subjects_file)
 
147
  for idx, sub_row in subjects_data.iterrows():
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
148
  yield idx, {
149
  "subject_id": sub_row["subject_id"],
150
  "treatment": sub_row["treatment"],
 
181
  "livingroom": byt_to_image(sub_row["image.livingroom"]),
182
  "bedroom": byt_to_image(sub_row["image.bedroom"])
183
  }
184
+ }
185
+