deep-patient-representation-mimiciii-multitask/HP.py at master · Yuqi92/deep-patient-representation-mimiciii-multitask · GitHub

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
# coding=utf-8

"""parameter used in the project"""


# mimic data
mimic_note_events = 'mimic_csv/NOTEEVENTS.csv'
mimic_admissions = 'mimic_csv/ADMISSIONS.csv'
mimic_patients = 'mimic_csv/PATIENTS.csv'

#result_csv = "/data/CNN_mimiciii_mortality_prediction/merged/file/result.csv"
data_directory = "merged/file/entire_file/"
insight_patient_data_directory = "merged/file_" + insight_patient_id + "/entire_file/"

def get_data_directory():
    if insight_patient_id is None:
        return data_directory
    else:
        return insight_patient_data_directory

patient_vector_directory = "merged/patient_vector/patient_vector_3_task/"

# if we set insight_patient_id to None, it means that we do not want to get insight of patient
insight_patient_id = "patient1231"
use_everything_to_test = True

result_csv = "merged/file/result_csv_dead_los.csv"
insight_patient_result_csv = "/merged/file_" + insight_patient_id + "/result_csv_dead_los.csv"


def get_result_csv():
    if insight_patient_id is None:
        return result_csv
    else:
        return insight_patient_result_csv


#subject_index = "/merged/file/subject.csv"

index_path = 'merged/index'

index_train_path = index_path + '/train.npy'
index_dev_path = index_path + '/dev.npy'
index_test_path = index_path + '/test.npy'
load_index = True
read_data_thread_num = 8

#pre-trained embedding
embedding_file = '/home/ysi/Documents/amia/cancer_ner_relation_v1/data/glove.6B/mimic.k100.w2v'

n_max_sentence_num = 1000 # truncated to 1000 sentences a document
n_max_word_num = 25 # truncated to 25 words a sentence

# note category
category = ['pad', 'Respiratory', 'ECG','Radiology','Nursing/other','Rehab Services','Nutrition','Pharmacy','Social Work',
            'Case Management','Physician','General','Nursing','Echo','Consult']
category_id = {cate: idx for idx, cate in enumerate(category)}


# mortality task
#single_task
tasks_dead_date = [366]
# 20-task
#tasks_dead_date = [0,5,14,31,43,68,103,142,196,269,366,453,573,711,893,1092,1342,1626,1997,2548]
# 5-task
#tasks_dead_date = [0, 31, 91, 366, 1095]
# 3-task
#tasks_dead_date = [0, 31, 366]

# length of stay
tasks_los_date = []

# CNN model hyperparameters
restore = True
multi_size = len(tasks_dead_date) + len(tasks_los_date)
embedding_size = 100
max_document_length = 1000
max_sentence_length = 25
n_batch = 64
early_stop_times = 5
filter_sizes = [3, 4, 5]
num_filters = 50
n_category = len(category)
dim_category = 10
document_filter_size = 3
document_num_filters = 50
learning_rate = 0.001
rate_decay = 0.99
drop_out_train = 0.8

model_type = "CNN"  # option: can be CNN or SIMPLE Feedforward NN

num_classes = 2
lambda_regularizer_strength = 5

model_folder = "merged/multi_task/result_3_task/model_1/"
model_path = model_folder+ "model.weights/model.ckpt"

# log
log_file_name = "example.log"

test_output = model_folder +'sentence_error_analysis_single_task_' + insight_patient_id + '.csv'

# tensorboard tsne visualization
tsne_vector_directory = 'merged/patient_vector/visualization/3_task/vector.npy'

tsne_label_directory = 'merged/patient_vector/visualization/3_task/label.txt'

tensorboard_log = 'project_tensorboard/log'