My commit
Browse files- dataprocessingcopy.py +203 -0
dataprocessingcopy.py
ADDED
@@ -0,0 +1,203 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# -*- coding: utf-8 -*-
|
2 |
+
"""dataprocessing.ipynb
|
3 |
+
|
4 |
+
Automatically generated by Colaboratory.
|
5 |
+
|
6 |
+
Original file is located at
|
7 |
+
https://colab.research.google.com/drive/10At7vh21OGTlE-Myv1NhAHi7l7NwBocQ
|
8 |
+
"""
|
9 |
+
|
10 |
+
import pandas as pd
|
11 |
+
import numpy as np
|
12 |
+
import os
|
13 |
+
from zipfile import ZipFile
|
14 |
+
import re
|
15 |
+
import json
|
16 |
+
import base64
|
17 |
+
|
18 |
+
from google.colab import drive
|
19 |
+
drive.mount('/content/drive')
|
20 |
+
|
21 |
+
path = "/content/drive/MyDrive/Duke/huggingface_project/data"
|
22 |
+
|
23 |
+
df = pd.read_excel(path+"/questionnaire-data.xlsx", header=2)
|
24 |
+
|
25 |
+
df["vviq_score"] = np.sum(df.filter(like = "vviq"), axis = 1)
|
26 |
+
df["osiq_score"] = np.sum(df.filter(like = "osiq"), axis = 1)
|
27 |
+
df["treatment"] = np.where(df.vviq_score > 40, "control", "aphantasia")
|
28 |
+
|
29 |
+
df = df.rename(columns={
|
30 |
+
"Sub ID": "sub_id",
|
31 |
+
df.columns[5]: "art_ability",
|
32 |
+
df.columns[6]: "art_experience",
|
33 |
+
df.columns[9]: "difficult",
|
34 |
+
df.columns[10]: "diff_explanation"
|
35 |
+
})
|
36 |
+
|
37 |
+
df.columns = df.columns.str.lower()
|
38 |
+
|
39 |
+
df = df.drop(df.filter(like="unnamed").columns, axis = 1)
|
40 |
+
|
41 |
+
df = df.drop(df.filter(regex="(vviq|osiq)\d+").columns, axis = 1)
|
42 |
+
|
43 |
+
df[df.columns[df.dtypes == "object"]] = df[df.columns[df.dtypes == "object"]].astype("string")
|
44 |
+
|
45 |
+
df[df.columns] = df[df.columns].replace([np.nan,pd.NA, "nan","na","NA","n/a","N/A","N/a"], None)
|
46 |
+
|
47 |
+
data = {}
|
48 |
+
for ind, row in df.iterrows():
|
49 |
+
data[row["sub_id"]] = {
|
50 |
+
"subject_id": int(row["sub_id"]),
|
51 |
+
"treatment": row["treatment"],
|
52 |
+
"demographics": dict(df.iloc[ind][1:-1])
|
53 |
+
}
|
54 |
+
data[row["sub_id"]]["demographics"]["art_ability"] = int(data[row["sub_id"]]["demographics"]["art_ability"])
|
55 |
+
data[row["sub_id"]]["demographics"]["vviq_score"] = int(data[row["sub_id"]]["demographics"]["vviq_score"])
|
56 |
+
data[row["sub_id"]]["demographics"]["osiq_score"] = int(data[row["sub_id"]]["demographics"]["osiq_score"])
|
57 |
+
|
58 |
+
stored_images = {}
|
59 |
+
with ZipFile(path + "/Images.zip", "r") as zip:
|
60 |
+
for image_file in zip.namelist():
|
61 |
+
with zip.open(image_file, 'r') as fil:
|
62 |
+
im = fil.read()
|
63 |
+
im_encoded = base64.b64encode(im).decode("utf-8")
|
64 |
+
stored_images[image_file.removesuffix(".jpg")] = im_encoded
|
65 |
+
|
66 |
+
def get_sub_files(subject, file_list):
|
67 |
+
pattern = re.compile("^.*" + subject + "-[a-z]{3}\d-(kitchen|livingroom|bedroom).*")
|
68 |
+
sub_files = [f for f in file_list if pattern.match(f)]
|
69 |
+
sub = {
|
70 |
+
"kitchen": {
|
71 |
+
"perception": "",
|
72 |
+
"memory": ""
|
73 |
+
},
|
74 |
+
"livingroom": {
|
75 |
+
"perception": "",
|
76 |
+
"memory": ""
|
77 |
+
},
|
78 |
+
"bedroom": {
|
79 |
+
"perception": "",
|
80 |
+
"memory": ""
|
81 |
+
},
|
82 |
+
}
|
83 |
+
|
84 |
+
for fil in sub_files:
|
85 |
+
if "kitchen" in fil:
|
86 |
+
if "pic" in fil:
|
87 |
+
sub["kitchen"]["perception"] = fil
|
88 |
+
else:
|
89 |
+
sub["kitchen"]["memory"] = fil
|
90 |
+
elif "livingroom" in fil:
|
91 |
+
if "pic" in fil:
|
92 |
+
sub["livingroom"]["perception"] = fil
|
93 |
+
else:
|
94 |
+
sub["livingroom"]["memory"] = fil
|
95 |
+
else:
|
96 |
+
if "pic" in fil:
|
97 |
+
sub["bedroom"]["perception"] = fil
|
98 |
+
else:
|
99 |
+
sub["bedroom"]["memory"] = fil
|
100 |
+
return sub
|
101 |
+
|
102 |
+
with ZipFile(path + "/Aphantasia-Drawings.zip", "r") as zip:
|
103 |
+
files = zip.namelist()
|
104 |
+
aphan_subs = list({f.split("/")[0] for f in files})
|
105 |
+
aphantasia_drawing_dataset = {}
|
106 |
+
for s in aphan_subs:
|
107 |
+
if int(s[3:]) in data.keys():
|
108 |
+
data[int(s[3:])]["drawings"] = get_sub_files(s, files)
|
109 |
+
else:
|
110 |
+
data[int(s[3:])] = {"drawings": get_sub_files(s,files)}
|
111 |
+
|
112 |
+
with ZipFile(path + "/Control-Drawings.zip", "r") as zip:
|
113 |
+
files = zip.namelist()
|
114 |
+
cntrl_subs = list({f.split("/")[0] for f in files})
|
115 |
+
full_control = {}
|
116 |
+
for s in cntrl_subs:
|
117 |
+
if int(s[3:]) in data.keys():
|
118 |
+
data[int(s[3:])]["drawings"] = get_sub_files(s, files)
|
119 |
+
else:
|
120 |
+
data[int(s[3:])] = {"drawings": get_sub_files(s,files)}
|
121 |
+
|
122 |
+
stored_images["kitchen"] = stored_images.pop('high_sun_ajwbpqrwvknlvpeh')
|
123 |
+
stored_images["bedroom"] = stored_images.pop('low_sun_acqsqjhtcbxeomux')
|
124 |
+
stored_images["livingroom"] = stored_images.pop('low_sun_byqgoskwpvsbllvy')
|
125 |
+
|
126 |
+
def extract_images(subject, treatment):
|
127 |
+
images_bytes = {
|
128 |
+
"kitchen": {
|
129 |
+
"perception": "",
|
130 |
+
"memory": ""
|
131 |
+
},
|
132 |
+
"livingroom": {
|
133 |
+
"perception": "",
|
134 |
+
"memory": ""
|
135 |
+
},
|
136 |
+
"bedroom": {
|
137 |
+
"perception": "",
|
138 |
+
"memory": ""
|
139 |
+
}
|
140 |
+
}
|
141 |
+
for room in ["kitchen", "livingroom", "bedroom"]:
|
142 |
+
paths = data[subject]["drawings"].get(room).values()
|
143 |
+
paths = [p for p in paths if p != ""]
|
144 |
+
if treatment == "aphantasia":
|
145 |
+
with ZipFile(path + "/Aphantasia-Drawings.zip", "r") as zip:
|
146 |
+
for filename in paths:
|
147 |
+
with zip.open(filename, 'r') as fil:
|
148 |
+
im = fil.read()
|
149 |
+
im_encoded = base64.b64encode(im).decode("utf-8")
|
150 |
+
if "mem" in filename:
|
151 |
+
images_bytes[room]["memory"] = im_encoded
|
152 |
+
else:
|
153 |
+
images_bytes[room]["perception"] = im_encoded
|
154 |
+
else:
|
155 |
+
with ZipFile(path + "/Control-Drawings.zip", "r") as zip:
|
156 |
+
for filename in paths:
|
157 |
+
with zip.open(filename, 'r') as fil:
|
158 |
+
im = fil.read()
|
159 |
+
im_encoded = base64.b64encode(im).decode("utf-8")
|
160 |
+
if "mem" in filename:
|
161 |
+
images_bytes[room]["memory"] = im_encoded
|
162 |
+
else:
|
163 |
+
images_bytes[room]["perception"] = im_encoded
|
164 |
+
|
165 |
+
return images_bytes
|
166 |
+
|
167 |
+
missing = []
|
168 |
+
for i in data.keys():
|
169 |
+
if "drawings" in data[i] and "treatment" in data[i]:
|
170 |
+
data[i]["drawings"] = extract_images(i,data[i]["treatment"])
|
171 |
+
else:
|
172 |
+
missing.append(i)
|
173 |
+
|
174 |
+
for num in missing:
|
175 |
+
data.pop(num, None)
|
176 |
+
|
177 |
+
for sub in data.keys():
|
178 |
+
data[sub]["image"] = stored_images
|
179 |
+
|
180 |
+
subject_data_path = path + "/clean_data.json"
|
181 |
+
|
182 |
+
#with open(subject_data_path, "w", encoding="utf-8") as sub_data:
|
183 |
+
#json.dump(data, sub_data, indent=2)
|
184 |
+
|
185 |
+
type(data)
|
186 |
+
|
187 |
+
da = pd.DataFrame(data)
|
188 |
+
|
189 |
+
|
190 |
+
|
191 |
+
|
192 |
+
|
193 |
+
flattened_data = []
|
194 |
+
|
195 |
+
for key, value in data.items():
|
196 |
+
flattened_subject = pd.json_normalize(value, sep='_')
|
197 |
+
flattened_data.append(flattened_subject)
|
198 |
+
|
199 |
+
|
200 |
+
da = pd.concat(flattened_data, ignore_index=True)
|
201 |
+
|
202 |
+
# Save the DataFrame to a Parquet file
|
203 |
+
da.to_parquet(path + 'data.parquet')
|