This repository has been archived by the owner on Apr 13, 2021. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcase_predictors.py
193 lines (140 loc) · 6.25 KB
/
case_predictors.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
185
186
187
188
189
190
191
192
193
import epispot as epi
import numpy as np # lgtm [py/unused-import]
from copy import deepcopy
def predict_short_term(case_data, entries, initial, end, n, time):
"""
Use this for short-term predictions only.
Long-term predictions may be inaccurate due to the prediction mechanism used.
Constructs an SIR Model to track the growth of cases given a predicted R Naught value and initial recovered
population from the data
:param case_data: Past case data organized earliest-latest (preferrably from the last 30 days)
:param entries: Number of entries in the case_data file
:param initial: Initial cases (at start of case_data)
:param end: Final number of cases (at end of case_data)
:param n: Total population of the region in question
:param time: Time for prediction (in days)
:return: Predicted cases
"""
def build_model(params):
"""
Builds a new epispot model and returns results.
Used for parameter fitting.
Several parameters have been pre-fitted from existing data.
"""
r_0 = params[0]
initial_recovered = params[1]
# Parameter Definitions
def N(t):
return n
def R_0(t):
return r_0
def gamma(t):
return 1 / 8.89375
def p_rec(t):
return 1.0
def rec_rate(t):
return 1 / 23
# Model Build
"""
S --> I --> R
"""
Susceptible = epi.comps.Susceptible(0, R_0, gamma, N)
Infected = epi.comps.Infected(1, N, R_0=R_0, gamma=gamma, p_recovery=p_rec, recovery_rate=rec_rate)
Recovered = epi.comps.Recovered(2, p_from_inf=p_rec, from_inf_rate=rec_rate)
# Model Compiler
Compiled_Model = epi.models.Model(N(0), layers=[Susceptible, Infected, Recovered],
layer_names=['Susceptible', 'Infected', 'Recovered'],
layer_map=[[Infected], [Recovered], []])
# Model Output
initial_vector = [n - initial - initial_recovered, initial, initial_recovered]
result = Compiled_Model.integrate(range(0, entries), starting_state=initial_vector)
formatted = []
for system in result:
formatted.append([deepcopy(system)[1]])
return formatted
def compile_model(params):
"""
Builds a new epispot model and returns results.
Used for parameter fitting.
Several parameters have been pre-fitted from existing data.
"""
r_0 = params[0]
# Parameter Definitions
def N(t):
return n
def R_0(t):
return r_0
def gamma(t):
return 1 / 8.89375
def p_rec(t):
return 1.0
def rec_rate(t):
return 1 / 23
# Model Build
"""
S --> I --> R
"""
Susceptible = epi.comps.Susceptible(0, R_0, gamma, N)
Infected = epi.comps.Infected(1, N, R_0=R_0, gamma=gamma, p_recovery=p_rec, recovery_rate=rec_rate)
Recovered = epi.comps.Recovered(2, p_from_inf=p_rec, from_inf_rate=rec_rate)
# Model Compiler
Compiled_Model = epi.models.Model(N(0), layers=[Susceptible, Infected, Recovered],
layer_names=['Susceptible', 'Infected', 'Recovered'],
layer_map=[[Infected], [Recovered], []])
return Compiled_Model
params_to_build = [0.007 * n, 2.0]
# optimized_parameters = epi.fitters.grad_des(build_model, case_data, params_to_build,
# 0.3, 5, n, range(entries))
ranges = [[0.0 + 0.25 * k for k in range(11)], [0.0 + 0.001 * n * k for k in range(22)]]
optimized_parameters = epi.fitters.tree_search(build_model, case_data, params_to_build, ranges, 3, n,
range(entries), verbose=False)
print('\nOptimization complete. A verbose log of the optimized parameters is shown below.')
print(optimized_parameters)
Model = compile_model(optimized_parameters)
print('\nModel compiled.')
R_0 = optimized_parameters[0] # lgtm [py/unused-local-variable]
initial_recovered = optimized_parameters[1]
predictions = Model.integrate(range(0, time),
starting_state=[n - end - initial_recovered, end, initial_recovered])
return predictions[-1][1]
def predict_uncontrolled(total_cases, current_cases, n, time):
"""
Predict the uncontrolled spread of COVID-19 over a set amount of time
:param total_cases: total confirmed COVID-19 cases
:param current_cases: number of currently infectious persons
:param n: total population
:param time: Time to predict (in days)
:return: cases after `time` days
"""
def N(t):
return n
def R_0(t):
return 2.5
def gamma(t):
return 1 / 8.89375
def p_rec(t):
return 1.0
def rec_rate(t):
return 1 / 23
# Model Build
"""
S --> I --> R
"""
Susceptible = epi.comps.Susceptible(0, R_0, gamma, N)
Infected = epi.comps.Infected(1, N, R_0=R_0, gamma=gamma, p_recovery=p_rec, recovery_rate=rec_rate)
Recovered = epi.comps.Recovered(2, p_from_inf=p_rec, from_inf_rate=rec_rate)
# Model Compiler
Compiled_Model = epi.models.Model(N(0), layers=[Susceptible, Infected, Recovered],
layer_names=['Susceptible', 'Infected', 'Recovered'],
layer_map=[[Infected], [Recovered], []])
result = Compiled_Model.integrate(range(0, time + 1),
starting_state=[n - total_cases, current_cases, total_cases - current_cases])
return round(result[-1][1]) # most recent, infected category
# predict short term
"""
data = open('test-data/sf-total-cases.csv') # .readlines()
# cases = [data[l].split(',')[1] for l in range(len(data))][-30:]
print(predict_short_term(data, 30, 2766, 4430, 883305, 7))
"""
# predict uncontrolled
# print(predict_uncontrolled(94315331, 24950778, 7e9, 1))