-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata_utils.py
184 lines (165 loc) · 9.42 KB
/
data_utils.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
from public_arts_app.models import Answers, Questions, InferenceOrder
from public_arts_app.models import PolicyStateActionPolicyStateAction, StateState, Questions, Answers, SurveyQuestions, GetnextquestionAskedquestion as gnq,\
AnswerquestionRespondedanswer as aqra, CreateparticipantParticipant as cpp, ParticipantSurveyQuestions as psq, SurveyAnswers, InferenceOrder, SpecialInferences
import pandas as pd
from public_arts_app.models import database
import pdb
def write_answers(df):
for idx, row in df.iterrows():
aid = row['answer_id']
a = Answers.get_or_none(answer_id = aid)
if not a:
continue
# if a.answer_text:
# continue
print(row['answer_id'])
print(row['answer_text'])
a.answer_text = row['answer_text']
a.save(only=[Answers.answer_text])
def write_questions(df):
#df = pd.read_csv("public_arts_app/datasets/final_annotations_transformed.csv")
for idx, row in df.iterrows():
sk = row['survey_key']
q = Questions.get_or_none(survey_key = sk)
if not q:
continue
q.infered_text = row['infered_text']
#q.inferedby_text = row['question_inferenceby']
#q.question_category = row['question type']
q.save(only=[Questions.infered_text])
def insert_inf_order():
group_size = 3
sequences = [
[1,2,3],
[1,3,2],
[3,1,2],
[3,2,1],
[2,3,1],
[2,1,3]
]
values = [(",".join([str(val) for val in s]),group_size) for s in sequences]
InferenceOrder.insert_many(values, fields = [InferenceOrder.inference_sequence, InferenceOrder.inference_group_size]).execute()
def write_answer_specials():
# excluded_vars = ["WKSWORKORG", "WRKOFFER", "STATECERT", "ACTSAME", "WHYUNEMP", "WNFTLOOK", "NILFACT", "UHRSWORKT", "EDDIPGED",
# "EDHGCGED", "DIFFHEAR", "DIFFEYE", "DIFFREM", "DIFFPHYS", "DIFFCARE", "DIFFMOB", "DIFFANY",
# "NMOTHERS", "NFATHERS", "AHRSWORKT", "EDHGCGED"]
# df = pd.read_csv("public_arts_app/datasets/special_answers.csv")
# special_answers = []
# for idx, row in df.iterrows():
# if row['do not infer'] == 0 or row['survey_key'] in excluded_vars:
# Questions.update({Questions.infer: 0}).where(Questions.question_id == row['question_id']).execute()
# if row['infer'] == 0 or row['do not infer'] == 0:
# Answers.update({Answers.infer: 0}).where(Answers.answer_id == row['answer_id']).execute()
# if row['infer'] == 1:
# special_answers.append((row['answer_id'], row['infered_text']))
#SpecialInferences.insert_many(special_answers, fields = [SpecialInferences.answer_id, SpecialInferences.inference_text]).execute()
pass
def get_study1_data():
study_num = '1'
qa = gnq.select(Questions.question_text.alias('question_text'),
Questions.survey_key.alias('survey_key'),
Answers.answer_text.alias('answer_text'),
aqra.answer_id.alias('answer_id'),
Answers.value.alias('answer_value'),
cpp.user_id.alias('user_id'),
cpp.uuid.alias('participant_id'),
cpp.done
).join(Questions, on=(
gnq.question_id==Questions.question_id)).switch(gnq).join(
aqra, on=(gnq.gnq_id==aqra.gnq_id)).join(
Answers, on=(aqra.answer_id==Answers.answer_id)).switch(gnq).join(
cpp, on=(gnq.user_id==cpp.user_id)
).where((cpp.study_num == study_num) & (cpp.done == 1)).dicts()
dfq = pd.DataFrame(list(qa))
user_ids = dfq['user_id'].unique().tolist()
qa_inference = psq.select(Questions.question_text.alias('question_text'),
Answers.answer_text.alias('answer_text'),
Questions.survey_key.alias('survey_key'),
psq.inference_answer_id.alias('answer_id'),
psq.inference_question_id,
psq.inference_question_category.alias('question_category'),
psq.answer_text.alias('share_decision'),
psq.inference_correct.alias('inference_correct'),
psq.user_id.alias('user_id')
).join(
Questions,on = (psq.inference_question_id==Questions.question_id)).switch(
psq).join(Answers, on=(psq.inference_answer_id==Answers.answer_id)).switch(
psq).join(cpp, on=(psq.user_id==cpp.user_id)).where(cpp.user_id.in_(user_ids)).dicts()
qa_survey = SurveyQuestions.select(
SurveyQuestions.question_text,
SurveyQuestions.question_options,
SurveyQuestions.sq_id,
SurveyAnswers.answer_text,
SurveyAnswers.user_id.alias('user_id')
).join(SurveyAnswers, on=(SurveyQuestions.sq_id==SurveyAnswers.sq_id)
).join(cpp, on=SurveyAnswers.user_id==cpp.user_id).where(cpp.user_id.in_(user_ids)).dicts()
df_inference = pd.DataFrame(list(qa_inference))
df_survey = pd.DataFrame(list(qa_survey))
user_ids = df_inference[~pd.isna(df_inference['inference_correct'])]['user_id'].unique()
dfq = dfq[dfq['user_id'].isin(user_ids)]
df_inference = df_inference[df_inference['user_id'].isin(user_ids)]
df_survey = df_survey[df_survey['user_id'].isin(user_ids)]
return dfq, df_inference, df_survey
def get_study2_data(for_analysis = False):
#cutoff = datetime(2023, 7, 26, 15, 0, 0, 0)
study_num = str(2)
database.connect(reuse_if_open=True)
qa = gnq.select(Questions.question_text.alias('question_text'),
Questions.survey_key.alias('survey_key'),
Answers.answer_text.alias('answer_text'),
aqra.answer_id.alias('answer_id'),
Answers.value.alias('answer_value'),
cpp.user_id.alias('user_id'),
cpp.uuid.alias('participant_id'),
cpp.user_condition,
cpp.done
).join(Questions, on=(
gnq.question_id==Questions.question_id)).switch(gnq).join(
aqra, on=(gnq.gnq_id==aqra.gnq_id)).join(
Answers, on=(aqra.answer_id==Answers.answer_id)).switch(gnq).join(
cpp, on=(gnq.user_id==cpp.user_id)
).where((cpp.study_num == study_num) & (cpp.done == 1)).dicts()
dfq = pd.DataFrame(list(qa))
user_ids = dfq['user_id'].unique().tolist()
qa_inference = psq.select(Questions.question_category.alias('inferenceby_category'),
Questions.question_text.alias('inferenceby_question_text'),
psq.inferedby_answer_id,
psq.inferedby_question_id,
psq.inference_question_id,
psq.inference_answer_id,
psq.inference_question_category.alias('inference_category'),
psq.answer_text.alias('share_decision'),
psq.inference_correct.alias('inference_correct'),
psq.user_id.alias('user_id'),
cpp.user_condition.alias('inference_condition') # 1,2,3,4
).join(
Questions,on = (psq.inferedby_question_id==Questions.question_id)).switch(
psq).join(cpp, on=(psq.user_id==cpp.user_id)).where(cpp.user_id.in_(user_ids)).dicts()
qdict = pd.DataFrame(Questions.select(Questions.question_id, Questions.question_text).dicts())
qa_survey = SurveyQuestions.select(
SurveyQuestions.question_text,
SurveyQuestions.question_options,
SurveyQuestions.sq_id,
SurveyAnswers.answer_text,
cpp.user_condition,
cpp.user_id
).join(SurveyAnswers, on=(SurveyQuestions.sq_id==SurveyAnswers.sq_id)
).join(cpp, on=SurveyAnswers.user_id==cpp.user_id).where(cpp.user_id.in_(user_ids)).dicts()
df_inference = pd.DataFrame(list(qa_inference))
df_inference['inference_question_text'] = df_inference['inference_question_id'].apply(lambda k: qdict[qdict['question_id'] == int(k)]['question_text'].iloc[0])
df_inference['inferenceby_question_text'] = df_inference['inferedby_question_id'].apply(lambda k: qdict[qdict['question_id'] == int(k)]['question_text'].iloc[0])
cat_map = {1: "arts", 2: "ads", 3: "protected", 4: "no_inference"}
df_inference['inference_category'] = df_inference['inference_category'].apply(lambda k: cat_map[k])
df_inference['inference_condition_name'] = df_inference['inference_condition'].apply(lambda k: cat_map[k])
df_survey = pd.DataFrame(list(qa_survey))
if for_analysis:
#df_inference.loc[df_inference['inference_condition'] == 4, 'inferenceby_category'] = 'Invalid'
df_inference.loc[df_inference['inference_condition'] == 4, 'inference_correct'] = 3
df_inference.loc[df_inference['inference_condition'] == 4, 'inference_condition'] = 0
df_inference.loc[df_inference['inference_correct'] == 3, 'inference_correct'] = 0
df_survey.loc[df_survey['user_condition'] == 4, 'user_condition'] = 0
database.close()
return dfq, df_inference, df_survey
# dfs = df_survey[df_survey['sq_id'] == 10]
# dfs['resp'] = dfs['answer_text'].apply(lambda k: re.findall("<b>([a-zA-Z ]+)</b>", k)[0])
# dfs.groupby('resp').size()