generated from opensafely/research-template
/
run_models.R
214 lines (161 loc) · 7.46 KB
/
run_models.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
################################################################################
# Description: Run all models on training data
#
# Author: Emily S Nightingale
# Date: 09/09/2020
#
################################################################################
library(tidyverse)
library(data.table)
library(sandwich)
library(boot)
library(lmtest)
library(sf)
# args <- c("analysisdata.rds", "community_prevalence.rds", "data/msoa_shp.rds", 0.1)
args <- commandArgs(trailingOnly=TRUE)
sink("./output_model_run.txt")
write("Run models",file="log_model_run.txt")
###############################################################################
dat <- readRDS(args[1])
tpp_cov <- readRDS(args[2])
msoa_shp <- readRDS(args[3])
test_sample <- as.numeric(args[4])
# --------------------------------- Subset data ------------------------------ #
# Remove rows with NA for any covariate of interest
dat_na_rm <- dat %>%
filter_at(vars(ch_size, ch_type, imd_quint, rural_urban, hh_med_age, hh_p_female, hh_dem_gt25, hh_prop_min), all_vars(!is.na(.)))
write(paste0("Rows excluded due to missing covariates: n = ",nrow(dat)-nrow(dat_na_rm)),file="log_model_run.txt", append = T)
dat <- dat_na_rm
# Subset time to account for 7-day time lag
dat <- dat %>%
filter_at(vars(probable_cases_rate,probable_chg7,probable_roll7,probable_roll7_lag2wk), all_vars(!is.na(.)))
print("Summary: NA filtered data")
summary(dat)
print("No. care homes in final analysis data:")
n_distinct(dat$household_id)
# Add 1 to probable cases to use log transform
dat %>%
mutate(across(c(probable_cases_rate, probable_roll7,probable_roll7_lag2wk,probable_roll7_lag2wk), function(x) log((x+1),base = 2), .names = "log2_{.col}")) -> dat
# ------------------------ Split data into training and test------------------ #
samp <- sample(unique(dat$household_id),(1-test_sample)*n_distinct(dat$household_id))
train <- filter(dat, household_id %in% samp)
test <- filter(dat, !household_id %in% samp)
# Save test data for use in model validation
saveRDS(test,"./testdata.rds")
print("No. care homes in training data:")
train %>%
group_by(ever_affected) %>%
summarise(n_ch = n_distinct(household_id))
print("No. care homes in testing data:")
test %>%
group_by(ever_affected, wave) %>%
summarise(n_ch = n_distinct(household_id))
## ----------------------------- Model Formulae -------------------------------##
# Baseline: static risk factors, no time-varying community risk
f0 <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + wave
# Time-varying (1): current day cases
f1 <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + wave + log2_probable_cases_rate
# Time-varying (2): 7-day change
f2 <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + wave + probable_chg7
# Time-varying (3): 7-day rolling average
f3 <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + wave + log2_probable_roll7
# Time varying (4): Lagged
f4a <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + wave + log2_probable_roll7_lag1wk
f4b <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + log2_probable_roll7_lag2wk
# # Time interaction (5)
f5a <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + log2_probable_roll7*wave
f5b <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + log2_probable_roll7_lag1wk*wave
f5c <- event_ahead ~ ch_size + ch_type + imd_quint + rural_urban + hh_med_age + hh_p_female + hh_dem_gt25 + hh_prop_min + log2_probable_roll7_lag2wk*wave
formulae <- list(base = f0, fixed = f1, week_change = f2, roll_avg = f3, roll_avg_lag1 = f4a, roll_avg_lag2 = f4b,
interaction = f5a, interaction_lag1 = f5b, interaction_lag2 = f5c)
# f00 <- event_ahead ~ 1
# f0 <- event_ahead ~ ch_size
# f1 <- event_ahead ~ ch_type
# f2 <- event_ahead ~ hh_med_age
# f3 <- event_ahead ~ hh_p_female
# f4 <- event_ahead ~ hh_p_dem
# f5 <- event_ahead ~ probable_cases_rate
# f6 <- event_ahead ~ probable_roll7
# f7 <- event_ahead ~ probable_roll7_lag1wk
# f8 <- event_ahead ~ probable_roll7_lag2wk
#
# formulae_test <- list(f00,f0,f1,f2,f3,f4,f5,f6,f7,f8)
## ------------------------- Check variable levels ---------------------------##
print("Summary: Training data")
summary(train)
## --------------------------------- Fitting --------------------------------- ##
fit_mods <- function(formulae){
out <- tryCatch(
{
message("Attempting model fit")
lapply(formulae, function(f) stats::glm(f, family = "binomial", data = train))
},
error=function(cond) {
message("Error in model fitting")
message(cond)
# Choose a return value in case of error
return(NA)
}
)
return(out)
}
time1 <- Sys.time()
fits <- fit_mods(formulae)
write(paste0("Time fitting models: ",round(time1-Sys.time(),2)), file="log_model_run.txt", append = TRUE)
print("Summary: Model fits")
lapply(fits, summary)
# ypred<-predict(fits[[7]], newdata = train, type = "response")
# plot(train$day,train$event_ahead)
# points(train$day,ypred,col="blue")
# Robust SEs for coefficient significance:
print_coeffs <- function(fit){
confints <- coefci(fit, df = Inf, vcov = vcovCL, cluster = train$household_id)
testcoeffs <- lmtest::coeftest(fit, vcov = vcovCL, cluster = train$household_id)
out <- as.data.frame(round(cbind(exp(cbind(testcoeffs[,1],confints,testcoeffs[,2])),testcoeffs[,3:4]),4))
names(out) <- c("Estimate","2.5%","97.5%","Std. Err.","z","Pr(>|z|)")
# print(fit$formula)
return(out)
}
print("Summary: Model coeffs with robust SEs")
lapply(fits, print_coeffs)
print("Brier score w/ training data")
brier <- function(mod) mean((mod$fitted.values - train$event_ahead)^2)
brier_score_train <- lapply(fits, brier)
brier_score_train
print("10-fold cross-validation")
time2 <- Sys.time()
cv_err <- lapply(formulae, function(f) boot::cv.glm(data = train, glmfit = stats::glm(f, family = "binomial", data = train), K = 10))
write(paste0("Time running cross-validation: ",round(time2-Sys.time(),2)), file="log_model_run.txt", append = TRUE)
print("Cross-validated estimate of prediction error [raw / adj for k-fold rather than LOO]:")
err <- lapply(cv_err, function(cv) cv$delta[2])
print(err)
# fit_opt_brier <- fits[[which.min(brier_score_train)]]
# fit_opt_cv <- fits[[which.min(err)]]
## MAP RESIDUALS ##
# map_resids <- function(fit){
#
# train %>%
# mutate(res = residuals(fit, type = "pearson")) %>%
# group_by(msoa) %>%
# summarise(mean_res = mean(res, na.rm = TRUE)) -> msoa_resids
#
# try(
# msoa_shp %>%
# full_join(msoa_resids, by = c("MSOA11CD" = "msoa")) %>%
# ggplot(aes(geometry = geometry, fill = mean_res)) +
# geom_sf(lwd = 0) +
# scale_fill_viridis_c() +
# theme_minimal() -> map, silent = TRUE)
#
# return(map)
#
#
# }
# pdf("model_resids_map.pdf", height = 10, width = 8)
# lapply(fits, map_resids)
# dev.off()
################################################################################
saveRDS(fits, "./fits.rds")
################################################################################
sink()
################################################################################