/
cpp_funcs.R
388 lines (353 loc) · 20.7 KB
/
cpp_funcs.R
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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
# For fitting
# Give names to elements of x from priors
named_params = function(priors)
{
paste(mapply(function(nm, i) paste0("double x_", nm, " = x[", i - 1, "]; (void) x_", nm, ";"), names(priors), seq_along(priors)), collapse = "\n");
}
# initialiser list with values from x
cpp_vec = function(x)
{
paste("{", paste(x, collapse = ", "), "}")
}
asc = function(x, y0, y1, s0, s1)
{
xx = s0 + x * (s1 - s0);
h0 = exp(s0) / (1 + exp(s0));
h1 = exp(s1) / (1 + exp(s1));
h = (exp(xx) / (1 + exp(xx)) - h0) / (h1 - h0);
return (y0 + (y1 - y0) * h);
}
# Wrap Cpp funcs for multi region fitting
wrap_region = function(model_v2, name, region_names)
{
ret = 'if (false) {}';
for (i in seq_along(region_names))
{
ret = c(ret,
glue::glue('else if (P.pop[0].name == "{region_names[i]}")'),
'{',
model_v2[[i]][[name]],
'}'
)
}
return (ret);
}
named_schedules = function()
{
"enum { CH_CONTACT, CH_TIER_2, CH_TIER_3, CH_CONTACT_CHANGE, CH_SEP_BOOST };"
}
# create cpp changes
cpp_chgI_voc = function(priors, seasonality, v2, v2_relu, v2_latdur, v2_serial, v2_infdur, v2_immesc, v2_ch_u)
{
glue::glue(
named_params(priors),
named_schedules(),
'vector<double> work_curve = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.008, 0.021, 0.033, 0.046, 0.058, 0.071, 0.083, 0.096, 0.108, 0.121, 0.133, 0.146, 0.158, 0.171, 0.183, 0.196, 0.208, 0.221, 0.233, 0.246, 0.258, 0.271, 0.283, 0.296, 0.308, 0.321, 0.334, 0.346, 0.359, 0.371, 0.384, 0.397, 0.41, 0.422, 0.435, 0.448, 0.461, 0.474, 0.487, 0.5, 0.513, 0.526, 0.539, 0.552, 0.566, 0.579, 0.592, 0.606, 0.619, 0.633, 0.646, 0.66, 0.674, 0.687, 0.701, 0.715, 0.729, 0.743, 0.757, 0.771, 0.785, 0.799, 0.813, 0.828, 0.842, 0.856, 0.87, 0.885, 0.899, 0.914, 0.928, 0.942, 0.957, 0.971, 0.986, 1, 1.014, 1.029, 1.043, 1.058, 1.072, 1.087, 1.101, 1.115, 1.13, 1.144, 1.159, 1.173, 1.188, 1.202, 1.216, 1.231, 1.245, 1.26, 1.274, 1.289, 1.303, 1.317, 1.332, 1.346, 1.361 };',
'vector<double> other_curve = { 0.064, 0.066, 0.067, 0.068, 0.069, 0.071, 0.072, 0.073, 0.075, 0.076, 0.077, 0.078, 0.08, 0.081, 0.082, 0.084, 0.085, 0.086, 0.087, 0.089, 0.09, 0.091, 0.092, 0.094, 0.095, 0.096, 0.098, 0.099, 0.1, 0.101, 0.103, 0.104, 0.105, 0.106, 0.108, 0.109, 0.11, 0.112, 0.113, 0.114, 0.116, 0.118, 0.119, 0.121, 0.123, 0.125, 0.128, 0.13, 0.132, 0.135, 0.137, 0.14, 0.143, 0.146, 0.15, 0.154, 0.159, 0.164, 0.169, 0.175, 0.182, 0.19, 0.198, 0.207, 0.217, 0.228, 0.24, 0.252, 0.266, 0.28, 0.295, 0.31, 0.327, 0.344, 0.361, 0.379, 0.398, 0.418, 0.438, 0.459, 0.48, 0.502, 0.525, 0.549, 0.572, 0.597, 0.621, 0.647, 0.672, 0.698, 0.725, 0.751, 0.778, 0.805, 0.833, 0.86, 0.888, 0.916, 0.944, 0.972, 1, 1.028, 1.056, 1.084, 1.112, 1.14, 1.168, 1.196, 1.224, 1.252, 1.28, 1.308, 1.337, 1.365, 1.393, 1.421, 1.449, 1.477, 1.505, 1.533, 1.561, 1.589, 1.617, 1.645, 1.673, 1.701 };',
'auto interp = [&](double y, vector<double>& curve) {',
' if (y < 0) return curve[0];',
' unsigned int i = (unsigned int)(y * 100);',
' if (i >= curve.size() - 1) return curve.back();',
' double f = y * 100 - i;',
' return (1 - f) * curve[i] + f * curve[i + 1];',
'};',
'auto odds = [&](double v, double lo) {',
' double a = v / (1 - v);',
' return a * exp(lo) / (a * exp(lo) + 1);',
'};',
if (seasonality) {
' P.pop[0].season_A[0] = x_seasonality;'
' P.pop[0].season_T[0] = 365.25;'
' P.pop[0].season_phi[0] = 0;'
} else '',
'for (unsigned int g = 0; g < P.processes[0].prob.size(); ++g) {',
' // To hospital',
' P.processes[0].prob[g][0] = odds(P.processes[0].prob[g][0], x_hosp_rlo);',
' P.processes[0].prob[g][1] = 1 - P.processes[0].prob[g][0];',
' P.processes[10].prob[g][0] = odds(P.processes[10].prob[g][0], x_hosp_rlo${ if (v2) "+ x_v2_hosp_rlo" else "" });',
' P.processes[10].prob[g][1] = 1 - P.processes[10].prob[g][0];',
' // To ICU',
' P.processes[1].prob[g][0] = odds(P.processes[1].prob[g][0], x_icu_rlo);',
' P.processes[1].prob[g][1] = 1 - P.processes[1].prob[g][0];',
' P.processes[11].prob[g][0] = odds(P.processes[11].prob[g][0], x_icu_rlo${ if (v2) "+ x_v2_icu_rlo" else "" });',
' P.processes[11].prob[g][1] = 1 - P.processes[11].prob[g][0];',
' // To death',
' P.processes[4].prob[g][0] = odds(P.processes[4].prob[g][0], x_cfr_rlo);',
' P.processes[4].prob[g][1] = 1 - P.processes[4].prob[g][0];',
' P.processes[14].prob[g][0] = odds(P.processes[14].prob[g][0], x_cfr_rlo${ if (v2) "+ x_v2_cfr_rlo" else "" });',
' P.processes[14].prob[g][1] = 1 - P.processes[14].prob[g][0];',
' // Relative susceptibility',
' P.pop[0].u[g] = P.pop[0].u[g] * x_u;',
if (v2_relu) {
' P.pop[0].u2[g] = P.pop[0].u2[g] * x_u * x_v2_relu;'
} else {
' P.pop[0].u2[g] = P.pop[0].u2[g] * x_u;'
},
if (v2_immesc) {
' P.pop[0].pi2_r[g] = x_v2_immesc;'
} else '',
'}',
if (v2_ch_u) {
'P.pop[0].u2[0] = P.pop[0].u2[0] * x_v2_ch_u;
P.pop[0].u2[1] = P.pop[0].u2[1] * x_v2_ch_u;
P.pop[0].u2[2] = P.pop[0].u2[2] * x_v2_ch_u;
P.pop[0].u2[3] = P.pop[0].u2[3] * x_v2_ch_u;'
} else '',
if (v2_latdur) {
'P.pop[0].dE2 = delay_gamma(x_v2_latdur * 2.5, 2.5, 15, 0.25);'
} else '',
if (v2_serial) {
'P.pop[0].dE2 = delay_gamma(x_v2_serial * 2.5, 2.5, 15, 0.25);
P.pop[0].dIp2 = delay_gamma(x_v2_serial * 2.5, 4.0, 30, 0.25);
P.pop[0].dIs2 = delay_gamma(x_v2_serial * 2.5, 4.0, 30, 0.25);
P.pop[0].dIa2 = delay_gamma(x_v2_serial * 5.0, 4.0, 30, 0.25);
for (unsigned int g = 0; g < P.pop[0].u2.size(); ++g) {
P.pop[0].u2[g] /= x_v2_serial;
}'
} else '',
if (v2_infdur) {
'P.pop[0].dIp2 = delay_gamma(x_v2_infdur * 2.5, 4.0, 30, 0.25);
P.pop[0].dIs2 = delay_gamma(x_v2_infdur * 2.5, 4.0, 30, 0.25);
P.pop[0].dIa2 = delay_gamma(x_v2_infdur * 5.0, 4.0, 30, 0.25);'
} else '',
'// Delays to admission to hospital & ICU',
'P.processes[0]. delays[0] = delay_gamma(x_hosp_admission, 0.71, 60, 0.25);',
'P.processes[10].delays[0] = delay_gamma(x_hosp_admission, 0.71, 60, 0.25);',
'P.processes[1]. delays[0] = delay_gamma(x_icu_admission, 1.91, 60, 0.25);',
'P.processes[11].delays[0] = delay_gamma(x_icu_admission, 1.91, 60, 0.25);',
'// Death',
'P.processes[4] .delays[0] = delay_gamma(x_death_mean, 2.2, 60, 0.25);',
'P.processes[14].delays[0] = delay_gamma(x_death_mean, 2.2, 60, 0.25);',
'// Seeding of original and variant strain',
'P.pop[0].seed_times = seq((int)x_tS, (int)x_tS + 27);',
if (v2) {
'P.pop[0].seed_times2 = vector<double>(10, x_v2_when);'
} else {
'P.pop[0].seed_times2 = vector<double>(1, 99999);'
},
'auto asc = [&](double x, double y0, double y1, double s0, double s1) {',
' double xx = s0 + x * (s1 - s0);',
' double h0 = exp(s0) / (1 + exp(s0));',
' double h1 = exp(s1) / (1 + exp(s1));',
' double h = (exp(xx) / (1 + exp(xx)) - h0) / (h1 - h0);',
' return y0 + (y1 - y0) * h;',
'};',
# Contact adjustment
'for (unsigned int i = 0; i < P.changes.ch[CH_CONTACT_CHANGE].times.size(); ++i) {',
' double tx = double(i) / (P.changes.ch[CH_CONTACT_CHANGE].times.size() - 1.0);',
' P.changes.ch[CH_CONTACT_CHANGE].values[i] = vector<double>(8, asc(tx, 1.0, x_contact_final, -x_contact_s0, x_contact_s1));',
'}',
# Sep boost
#'P.changes.ch[CH_SEP_BOOST].values[0] = vector<double>(8, x_sep_boost);',
#'P.changes.ch[CH_SEP_BOOST].times[0] = x_sep_when;',
# fitting of google mobility indices
'for (unsigned int k : vector<unsigned int> { CH_CONTACT, CH_TIER_2, CH_TIER_3 }) {',
' for (unsigned int i = 0; i < P.changes.ch[k].times.size(); ++i) {',
' //double resi = P.changes.ch[k].values[i][0];',
' double wplc = P.changes.ch[k].values[i][1];',
' double groc = P.changes.ch[k].values[i][2];',
' double rtrc = P.changes.ch[k].values[i][3];',
' double trns = P.changes.ch[k].values[i][4];',
' double scho = P.changes.ch[k].values[i][5];',
' double othx = rtrc * 0.345 + trns * 0.445 + groc * 0.210;',
' double t = P.changes.ch[k].times[i];',
# from CoMix analysis
' double home = asc(min(1.0, t / 365.0), 1.0, 1.545019 / 3.875622, -79 * 0.6, 286 * 0.6);',
' double work = interp(wplc, work_curve);',
' double othe = interp(othx, other_curve);',
' P.changes.ch[k].values[i] = { home, work, scho, othe, home, work, scho, othe };',
' }',
'}',
.sep = "\n ", .open = "${", .close = "}"
)
}
# create c++ likelihood components
cpp_likI_voc = function(params, ld, sitreps, sero, virus, sgtfd, popid, max_date, priors, death_cutoff, use_sgtf)
{
refdate = ld[, max(date)] + 1;
ld = copy(ld[date <= max_date]);
sitreps = copy(sitreps[date <= max_date]);
sero = copy(sero[End.date <= max_date]);
virus = copy(virus[End.date <= max_date]);
sgtfd = copy(sgtfd[date <= max_date]);
glue::glue(
named_params(priors),
named_schedules(),
'std::vector<double> death_v = ${ cpp_vec(ld[!is.na(N), N]) };',
'std::vector<double> death_t = ${ cpp_vec(ld[!is.na(N), as.numeric(ymd(date) - ymd(params$date0))]) };',
'std::vector<double> death_tr = ${ cpp_vec(ld[!is.na(N), as.numeric(refdate - ymd(date))]) };',
'std::vector<double> hosp_i_v = ${ cpp_vec(sitreps[!is.na(n_admitted_diagnosed), n_admitted_diagnosed]) };',
'std::vector<double> hosp_i_t = ${ cpp_vec(sitreps[!is.na(n_admitted_diagnosed), as.numeric(ymd(date) - ymd(params$date0))]) };',
'std::vector<double> hosp_p_v = ${ cpp_vec(sitreps[!is.na(n_in_all_beds), n_in_all_beds]) };',
'std::vector<double> hosp_p_t = ${ cpp_vec(sitreps[!is.na(n_in_all_beds), as.numeric(ymd(date) - ymd(params$date0))]) };',
'std::vector<double> icu_p_v = ${ cpp_vec(sitreps[!is.na(n_in_itu), n_in_itu]) };',
'std::vector<double> icu_p_t = ${ cpp_vec(sitreps[!is.na(n_in_itu), as.numeric(ymd(date) - ymd(params$date0))]) };',
'std::vector<double> sero_xi = ${ cpp_vec(sero[, xi]) };',
'std::vector<double> sero_om = ${ cpp_vec(sero[, omega]) };',
'std::vector<double> sero_al = ${ cpp_vec(sero[, alpha]) };',
'std::vector<double> sero_t0 = ${ cpp_vec(sero[, as.numeric(ymd(Start.date) - ymd(params$date0))]) };',
'std::vector<double> sero_t1 = ${ cpp_vec(sero[, as.numeric(ymd(End.date) - ymd(params$date0))]) };',
'std::vector<double> virus_xi = ${ cpp_vec(virus[, xi]) };',
'std::vector<double> virus_om = ${ cpp_vec(virus[, omega]) };',
'std::vector<double> virus_al = ${ cpp_vec(virus[, alpha]) };',
'std::vector<double> virus_t0 = ${ cpp_vec(virus[, as.numeric(ymd(Start.date) - ymd(params$date0))]) };',
'std::vector<double> virus_t1 = ${ cpp_vec(virus[, as.numeric(ymd(End.date) - ymd(params$date0))]) };',
if (use_sgtf) {
'std::vector<double> sgtf_s = ${ cpp_vec(sgtfd[!is.na(sgtf), sgtf]) };
std::vector<double> sgtf_o = ${ cpp_vec(sgtfd[!is.na(other), other]) };
std::vector<double> sgtf_t = ${ cpp_vec(sgtfd[!is.na(sgtf), as.numeric(ymd(date) - ymd(params$date0))]) };'
} else '',
'auto size_param = [](double disp) { return 1.0 / (disp * disp); };',
'for (int i = 0; i < int(death_t.size()) - ${death_cutoff}; ++i)',
' ll += nbinom(death_v[i], max(0.1, dyn("death_o", death_t[i], {}, {}) + dyn("death2_o", death_t[i], {}, {})), size_param(x_disp_deaths));',
'for (unsigned int i = 0; i < hosp_i_t.size(); ++i)',
' ll += nbinom(hosp_i_v[i], max(0.1, dyn("hosp_undetected_o", hosp_i_t[i], {}, {}) + dyn("hosp_undetected2_o", hosp_i_t[i], {}, {})), size_param(x_disp_hosp_inc));',
'for (unsigned int i = 0; i < hosp_p_t.size(); ++i)',
' ll += nbinom(hosp_p_v[i], max(0.1, dyn("hosp_p", hosp_p_t[i], {}, {}) + dyn("hosp2_p", hosp_p_t[i], {}, {}) - dyn("hosp_undetected_p", hosp_p_t[i], {}, {}) - dyn("hosp_undetected2_p", hosp_p_t[i], {}, {})), size_param(x_disp_hosp_prev));',
'for (unsigned int i = 0; i < icu_p_t.size(); ++i)',
' ll += nbinom(icu_p_v[i], max(0.1, dyn("icu_p", icu_p_t[i], {}, {}) + dyn("icu2_p", icu_p_t[i], {}, {})), size_param(x_disp_icu_prev));',
'for (unsigned int i = 0; i < virus_t0.size(); ++i) {',
' double virus_prev = 0;',
' for (double tt = virus_t0[i]; tt <= virus_t1[i]; ++tt)',
' virus_prev += dyn("pcr_positive_p", tt, {}, {}) / (virus_t1[i] - virus_t0[i] + 1);',
' ll += 10 * skewnorm(virus_prev / ${ sum(params$pop[[1]]$size) }, virus_xi[i], virus_om[i], virus_al[i]);',
'}',
'for (unsigned int i = 0; i < sero_t0.size(); ++i) {',
' double sero_prev = 0;',
' for (double tt = sero_t0[i]; tt <= sero_t1[i]; ++tt)',
' sero_prev += dyn("lfia_positive_p", tt, {}, {}) / (sero_t1[i] - sero_t0[i] + 1);',
' ll += skewnorm(sero_prev / ${ sum(params$pop[[1]]$size) }, sero_xi[i], sero_om[i], sero_al[i]);',
'}',
if (use_sgtf) {
'for (unsigned int i = 0; i < sgtf_t.size(); ++i) {
double s1 = dyn("test_o", sgtf_t[i], {}, {});
double s2 = dyn("test2_o", sgtf_t[i], {}, {});
double p2 = (s1 + s2) > 0 ? s2 / (s1 + s2) : 0;
double model_sgtf = p2 + (1 - p2) * x_v2_sgtf0;
ll += 10 * bbinom(sgtf_s[i], sgtf_s[i] + sgtf_o[i], model_sgtf, size_param(x_v2_disp));
}'
} else '',
.sep = "\n ", .open = "${", .close = "}"
)
}
# create c++ observer function
cpp_obsI_voc = function(concentration, v2, P.death, P.critical, priors)
{
glue::glue(
named_params(priors),
named_schedules(),
'auto asc = [&](double x, double y0, double y1, double s0, double s1) {',
' double xx = s0 + x * (s1 - s0);',
' double h0 = exp(s0) / (1 + exp(s0));',
' double h1 = exp(s1) / (1 + exp(s1));',
' double h = (exp(xx) / (1 + exp(xx)) - h0) / (h1 - h0);',
' return y0 + (y1 - y0) * h;',
'};',
'auto odds = [&](double v, double lo) {',
' double a = v / (1 - v);',
' return a * exp(lo) / (a * exp(lo) + 1);',
'};',
'auto clamp = [&](double v) {',
' return max(0.0, min(1.0, v));',
'};',
'dyn.Obs(t, 0, 0, 0) = estimate_Rt(P, dyn, t, 0, 50);',
'dyn.Obs(t, 0, 3, 0) = estimate_R0(P, dyn, t, 0, 50);',
'dyn.Obs(t, 0, 10, 0) = estimate_Rt_1(P, dyn, t, 0, 50);',
'dyn.Obs(t, 0, 11, 0) = estimate_Rt_2(P, dyn, t, 0, 50);',
# increase in young person mobility
if (concentration) {
'if (t == 182) {
double mode = 0.2;
double conc = x_concentration1;
double constant = 0.2;
for (unsigned int a = 0; a < P.pop[0].u.size(); ++a) {
P.pop[0].u[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
P.pop[0].u2[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
}
}
if (t == 213) {
double mode = 0.2;
double conc_prev = x_concentration1;
double conc = x_concentration2;
double constant = 0.2;
for (unsigned int a = 0; a < P.pop[0].u.size(); ++a) {
P.pop[0].u[a] /= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc_prev - 2) + 1, (1 - mode) * (conc_prev - 2) + 1) + constant;
P.pop[0].u[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
P.pop[0].u2[a] /= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc_prev - 2) + 1, (1 - mode) * (conc_prev - 2) + 1) + constant;
P.pop[0].u2[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
}
}
if (t == 244) {
double mode = 0.2;
double conc_prev = x_concentration2;
double conc = x_concentration3;
double constant = 0.2;
for (unsigned int a = 0; a < P.pop[0].u.size(); ++a) {
P.pop[0].u[a] /= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc_prev - 2) + 1, (1 - mode) * (conc_prev - 2) + 1) + constant;
P.pop[0].u[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
P.pop[0].u2[a] /= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc_prev - 2) + 1, (1 - mode) * (conc_prev - 2) + 1) + constant;
P.pop[0].u2[a] *= (1 - constant) * dbeta((a + 0.5) / P.pop[0].u.size(), mode * (conc - 2) + 1, (1 - mode) * (conc - 2) + 1) + constant;
}
}'
} else '',
# changing CFR and ICU admission
'if ((int)t % 7 == 0) {',
' double ifr_table[] = ${ cpp_vec(P.death) };',
' double icr_table[] = ${ cpp_vec(P.critical) };',
' double adj_f1 = asc(clamp(t / 190.0), 1.0, 0.0, -4.0, 1.0);',
' double adj_f3 = asc(clamp((t - 240.0) / 90.0), 0.0, 1.0, -5.0, 5.0);',
' double adj_f2 = 1.0 - adj_f1 - adj_f3;',
' double adj_c = asc(clamp(t/366.0), 0.0, 1.0, -6, 6);',
' for (unsigned int g = 0; g < P.processes[0].prob.size(); ++g) {',
' double ifr = ifr_table[g];',
' double icr = icr_table[g];',
' P.processes[4].prob[g][0] = odds(ifr, adj_f1 * x_cfr_rlo + adj_f2 * x_cfr_rlo2 + adj_f3 * x_cfr_rlo3);',
' P.processes[4].prob[g][1] = 1 - P.processes[4].prob[g][0];',
' P.processes[14].prob[g][0] = odds(ifr, adj_f1 * x_cfr_rlo + adj_f2 * x_cfr_rlo2 + adj_f3 * x_cfr_rlo3${ if (v2) "+ x_v2_cfr_rlo" else "" });',
' P.processes[14].prob[g][1] = 1 - P.processes[14].prob[g][0];',
' P.processes[1].prob[g][0] = odds(icr, (1 - adj_c) * x_icu_rlo + adj_c * x_icu_rlo2);',
' P.processes[1].prob[g][1] = 1 - P.processes[1].prob[g][0];',
' P.processes[11].prob[g][0] = odds(icr, (1 - adj_c) * x_icu_rlo + adj_c * x_icu_rlo2${ if (v2) "+ x_v2_icu_rlo" else "" });',
' P.processes[11].prob[g][1] = 1 - P.processes[11].prob[g][0];',
' }',
'}',
# changing detection rate in patients admitted to hospital
'double detection = asc(min(t / 365.0, 1.0), 14, 1, -5.86, 33.4);',
'P.processes[7].delays[0] = delay_gamma(detection, 0.59, 60, 0.25);',
'P.processes[15].delays[0] = delay_gamma(detection, 0.59, 60, 0.25);',
.sep = "\n ", .open = "${", .close = "}"
)
}
# Observer for vaccination programmes
cpp_obsI_vax = function(params, vacc)
{
glue::glue(
'if (t == 0) { dyn.scratch["vaxphase"] = 0; }',
'{',
'vector<double> vt = ${ cpp_vec(as.numeric(ymd(vacc$vt) - ymd(params$date0))) };',
'vector<double> v = ${ cpp_vec(unlist(vacc$v)) };',
'unsigned int phase = dyn.scratch["vaxphase"];',
'if (vt.size() > phase && t >= vt[phase]) {',
' int age_groups = P.pop[0].size.size();',
' P.pop[0].v = vector<double>(v.begin() + age_groups * phase, v.begin() + age_groups * (phase + 1));',
' dyn.scratch["vaxphase"] = phase + 1;',
'}',
'dyn.Obs(t, 0, 5, 0) = dyn.scratch["vaxphase"];',
'}',
.sep = "\n", .open = "${", .close = "}")
}
# Observer for forced seasonality
cpp_obsI_seasonality = function(forced_seasonality, seas_start_t)
{
glue::glue(
'if (t == ${seas_start_t}) {',
' for (unsigned int a = 0; a < P.pop[0].u.size(); ++a) {',
' P.pop[0].u[a] /= ${1 + forced_seasonality};',
' P.pop[0].u2[a] /= ${1 + forced_seasonality};',
' }',
' P.pop[0].season_A[0] = ${forced_seasonality};',
' P.pop[0].season_T[0] = 365.25;',
' P.pop[0].season_phi[0] = 0;',
'}',
.sep = "\n", .open = "${", .close = "}")
}