-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdiscretise_normalise.py
168 lines (116 loc) · 5.87 KB
/
discretise_normalise.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
import pandas as pd
import numpy as np
from pandas import DataFrame
from ED_preprocessing import generate_csv
NON_DISCRETISED_FILE_PATH = "GeneratedData/ED.csv"
numeric_columns = ['n_ed_visits',
'n_ed_admissions',
'triage_temp',
'triage_heartrate',
'triage_resprate',
'triage_o2sat',
'triage_sbp',
'triage_dbp']
# Discretise patient age into group
# 18-25, 26-45, 46-65, 66-85 and 85+
def discretise_age(dataframe: DataFrame) -> DataFrame:
print("Discretising age.....\n")
bins = [17, 25, 45, 65, 85, float('inf')]
labels = ['18-25', '26-45', '46-65', '66-85', '85+']
# Discretise the age column
dataframe['age_group'] = pd.cut(dataframe['age'], bins=bins, labels=labels, right=True)
return dataframe
# def discretise_temperature(dataframe: DataFrame) -> DataFrame:
# print("Discretising temperature.....\n")
# # Discretise the patient’s temperature in degrees Farenheit.
# bins = [0, 95, 96.8, 100.4, 102.2, float('inf')]
# labels = ['≤95', '95.1-96.8', '96.9-100.4', '100.5-102.2', '≥102.2']
# dataframe['temperature_group'] = pd.cut(dataframe['temperature'], bins=bins, labels=labels, right=True)
# return dataframe
# def discretise_heartrate(dataframe: DataFrame) -> DataFrame:
# print("Discretising heart rate.....\n")
# bins = [0, 40, 50, 90, 110, 130, float('inf')]
# labels = ['≤40', '41-50', '51-90', '91-110', '111-130', '≥131']
# dataframe['heartrate_group'] = pd.cut(dataframe['heartrate'], bins=bins, labels=labels, right=True)
# return dataframe
# def discretise_resprate(dataframe: DataFrame) -> DataFrame:
# print("Discretising respiratory rate.....\n")
# bins = [0, 8, 11, 20, 24, float('inf')]
# labels = ['≤8', '9-11', '12-20', '21-24', '≥25']
# dataframe['resrate_group'] = pd.cut(dataframe['resrate'], bins=bins, labels=labels, right=True)
# return dataframe
# def discretise_systolic_bp(dataframe: DataFrame) -> DataFrame:
# print("Discretising systolic blood pressure.....\n")
# bins = [0, 90, 100, 110, 219, float('inf')]
# labels = ['≤90', '91-100', '101-110', '111-219', '≥220']
# dataframe['sbp_group'] = pd.cut(dataframe['sbp'], bins=bins, labels=labels, right=True)
# return dataframe
# def discretise_o2sat(dataframe: DataFrame) -> DataFrame:
# print("Discretising oxygen saturation.....\n")
# bins = [0, 91, 93, 95, float('inf')]
# labels = ['≤91%', '92-93%', '94-95%', '≥96']
# dataframe['o2sat_group'] = pd.cut(dataframe['o2sat'], bins=bins, labels=labels, right=True)
# return dataframe
def discretise_presentation_time(df: DataFrame) -> DataFrame:
print("Discretising presentation time.....\n")
conditions = [
(df['presentation_hour'] >= 8) & (df['presentation_hour'] < 17), # 8am-5pm
(df['presentation_hour'] >= 17) & (df['presentation_hour'] < 23), # 5pm-11pm
(df['presentation_hour'] >= 23) | (df['presentation_hour'] < 8) # 11pm-8am
]
labels = ['business hours', 'evening', 'night']
df['presentation_time'] = np.select(conditions, labels, default='unknown')
return df
def discretise_LOS(dataframe: DataFrame) -> DataFrame:
print("Discretising ED Length of stay.....\n")
# Contains categories in 0-4, 5-12, 13-24 and 24+ measured in hours, as per RPA staff’s request.
bins = [0, 4, 12, 24, float('inf')]
# labels = ["0-4'", "'5-12'", "'13-24'", "'24+'"]
labels = ["0-4", "5-12", "13-24", "24+"]
dataframe['ED_LOS'] = pd.cut(dataframe['LOS (hours)'], bins=bins, labels=labels, right=True)
return dataframe
def discretise_triage_category(dataframe: DataFrame) -> DataFrame:
print("Converting triage category to nominal.....\n")
bins = [0, 1, 2, 3, 4, 5]
labels = ["one", "two", "three", "four", "five"]
dataframe['triage_category'] = pd.cut(dataframe['acuity'], bins=bins, labels=labels, right=True)
return dataframe
def discretise_pain_category(dataframe: DataFrame) -> DataFrame:
print("Converting triage category to nominal.....\n")
bins = [-1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
labels = ["zero", "one", "two", "three", "four", "five", "six", "seven", "eight", "nine", "ten"]
dataframe['triage_pain_discretised'] = pd.cut(dataframe['triage_pain'], bins=bins, labels=labels, right=True)
# dataframe['last_pain_discretised'] = pd.cut(dataframe['last_pain'], bins=bins, labels=labels, right=True)
return dataframe
def normalise(df: DataFrame) -> DataFrame:
print("Normalising numerical data.....\n")
for col in numeric_columns:
print("Normalising '" + col + "'column")
raw_name = 'raw_' + col
df.rename(columns={col: raw_name}, inplace=True)
df[col] = (df[raw_name] - df[raw_name].min()) / (df[raw_name].max() - df[raw_name].min())
return df
def main():
non_discretised_df = pd.read_csv(NON_DISCRETISED_FILE_PATH)
discretised_set = discretise_age(non_discretised_df)
discretised_set = discretise_presentation_time(discretised_set)
discretised_set = discretise_LOS(discretised_set)
discretised_set = discretise_triage_category(discretised_set)
discretised_set = discretise_pain_category(discretised_set)
discretised_set = normalise(discretised_set)
# Make the re_admitted column the last column again
cols = [col for col in discretised_set.columns if col != 'revisited'] + ['revisited']
discretised_set = discretised_set[cols]
generate_csv(discretised_set, 'GeneratedData/debug_discritised_ED.csv')
discretised_set = discretised_set.drop(columns=['age', 'LOS (hours)', 'acuity', 'presentation_hour', 'triage_pain'])
for col in numeric_columns:
raw_col = "raw_" + col
discretised_set = discretised_set.drop(columns=[raw_col])
discretised_set.rename(columns={'age_group': 'age'}, inplace=True)
discretised_set.rename(columns={'triage_pain_discretised': 'triage_pain'}, inplace=True)
generate_csv(discretised_set, 'GeneratedData/fully_processed_ED.csv')
# unique_values = discretised_set['diagnosis_category'].value_counts().count()
# print(unique_values)
# generate_csv(discretised_set.head(50000), 'GeneratedData/crop_discritised_ED.csv')
if __name__ == "__main__":
main()