/
run-model.R
51 lines (45 loc) · 2.09 KB
/
run-model.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
library(dplyr)
library(cmdstanr)
d <- read.csv('input/flu.csv') %>%
filter(female == 1) %>%
mutate(age = (age - 65)/10,
intercept = 1) %>%
rename(heart = heart.disease,
Z = treatment.assigned,
W_obs = treatment.received,
Y_obs = outcome)
d_00 <- d %>% filter(Z == 0, W_obs == 0)
d_01 <- d %>% filter(Z == 0, W_obs == 1)
d_10 <- d %>% filter(Z == 1, W_obs == 0)
d_11 <- d %>% filter(Z == 1, W_obs == 1)
X <- d %>% select(intercept, age, copd, heart)
X_00 <- d_00 %>% select(intercept, age, copd, heart)
X_01 <- d_01 %>% select(intercept, age, copd, heart)
X_10 <- d_10 %>% select(intercept, age, copd, heart)
X_11 <- d_11 %>% select(intercept, age, copd, heart)
data <- list(N = nrow(d), K = ncol(X_00),
N_00 = nrow(d_00), N_01 = nrow(d_01),
N_10 = nrow(d_10), N_11 = nrow(d_11),
Y_obs_00 = d_00$Y_obs, Y_obs_01 = d_01$Y_obs,
Y_obs_10 = d_10$Y_obs, Y_obs_11 = d_11$Y_obs,
X = X, X_00 = X_00, X_01 = X_01, X_10 = X_10, X_11 = X_11)
model <- cmdstan_model(stan_file = 'model/model.stan')
# setting of initial values
fit_MAP <- model$optimize(data = data, seed = 1, iter = 10000)
beta_co_c_ini <- fit_MAP$draws(variables = 'beta_co_c') %>% as.vector()
beta_co_t_ini <- fit_MAP$draws(variables = 'beta_co_t') %>% as.vector()
beta_nt_ini <- fit_MAP$draws(variables = 'beta_nt') %>% as.vector()
beta_at_ini <- fit_MAP$draws(variables = 'beta_at') %>% as.vector()
gamma_nt_ini <- fit_MAP$draws(variables = 'gamma_nt') %>% as.vector()
gamma_at_ini <- fit_MAP$draws(variables = 'gamma_at') %>% as.vector()
init_fun <- function(chain_id) {
set.seed(chain_id)
list(beta_co_c=beta_co_c_ini, beta_co_t=beta_co_t_ini,
beta_nt=beta_nt_ini, beta_at=beta_at_ini,
gamma_nt=gamma_nt_ini, gamma_at=gamma_at_ini)
}
fit <- model$sample(data = data, seed = 123, parallel_chains=4, init=init_fun,
iter_sampling = 2000, iter_warmup = 1000)
fit$save_object(file='output/result-model.RDS')
write.table(fit$summary(), file='output/fit-summary.csv',
sep=',', quote=TRUE, row.names=FALSE)