-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathr2r_forward_analysis.R
197 lines (179 loc) · 7.95 KB
/
r2r_forward_analysis.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
pb <- progress::progress_bar$new(
format = "[:bar] :current/:total (:percent in :elapsed) eta: :eta",
total = nrow(res_tab),
width = 80)
source("r2r_aicbic_select_forward.R")
# forward selection
for(i in 1:nrow(res_tab_forward)){
data_tmp <- gen_regress_data(scen = res_tab_forward$scen[i],
max_date = res_tab_forward$max_date[i],
reg = regions_dic)
res_tab_forward$select[i] <- list(aicbic_select_forward(optim_lag = res_tab_forward$optim_lag[i],
data = data_tmp))
pb$tick()
}
#
#
# lapply(res_tab_forward$select,"[[",3)[[1]]
# lapply(res_tab$select,"[[",3)[[1]]
lapply(res_tab_forward$select,"[[",3) %>%
map(mutate, rk_AIC = rank(AIC_val)) %>%
map(mutate, rk_BIC = rank(BIC_val, ties.method = "first")) %>%
map(filter, rk_AIC == 1| rk_BIC == 1) %>%
map(dplyr::select, model, rk_AIC, rk_BIC) %>%
map(pivot_longer, cols =starts_with ("rk"), names_to = "criterion") #%>%
map(filter, value == 1) %>%
bind_rows(.id = "set") %>%
dplyr::select(-value) %>%
mutate(criterion = gsub("rk_","",criterion)) -> chosen_forward
#
res_tab$select[[1]]$BIC_back$model[[9]]
res_tab_forward$select[[1]]$BIC_forward$model[[9]]
res_tab_forward[1:3] %>%
mutate(set = 1:n() %>%
as.character) %>%
left_join(chosen_forward, by = "set") %>%
filter(scen != "Mid") -> p_table_forward
p_val_forward <- list()
for(i in p_table_forward$set %>% unique){
p_table_forward %>%
arrange(set, criterion) %>%
filter(set == i) %>%
pull(model) -> model_selected
sum_AIC <- summary(res_tab_forward$select[[as.numeric(i)]][["AIC_forward"]]$model[model_selected[1]][[1]])
sum_AIC$coefficients[,4] %>%
enframe %>%
setNames(c("var","p_val")) %>%
mutate(p_lab = case_when(p_val > 0.05 ~ "ns",
p_val <= 0.05 & p_val > 0.01 ~ "*",
p_val <= 0.01 & p_val > 0.001 ~ "**",
p_val <= 0.001 ~ "***"),
set = i,
criterion = "AIC") -> res_1
sum_BIC <- summary(res_tab_forward$select[[as.numeric(i)]][["BIC_forward"]]$model[model_selected[2]][[1]])
sum_BIC$coefficients[,4] %>%
enframe %>%
setNames(c("var","p_val")) %>%
mutate(p_lab = case_when(p_val > 0.05 ~ "ns",
p_val <= 0.05 & p_val > 0.01 ~ "*",
p_val <= 0.01 & p_val > 0.001 ~ "**",
p_val <= 0.001 ~ "***"),
set = i,
criterion = "BIC") -> res_2
p_val_forward[[i]] <- rbind(res_1, res_2)
}
p_val_forward %<>% bind_rows()
#
lapply(lapply(res_tab_forward$select,"[[",1),"[[","var_combo") -> tmp
lapply(1:18, function(x) tmp[[x]] %>% do.call("rbind", .)) %>%
map(data.frame) %>%
map(mutate, model = 1:n()) %>%
bind_rows(.id = "set") -> AIC_panel_forward
lapply(lapply(res_tab_forward$select,"[[",2),"[[","var_combo") -> tmp
lapply(1:18, function(x) tmp[[x]] %>% do.call("rbind", .)) %>%
map(data.frame) %>%
map(mutate, model = 1:n()) %>%
bind_rows(.id = "set") -> BIC_panel_forward
#
chosen_forward[chosen_forward$criterion=="AIC",] %>%
left_join(AIC_panel_forward, by = c("set","model")) %>%
pivot_longer(cols = starts_with("X"), names_to = "var") %>%
bind_rows(chosen_forward[chosen_forward$criterion=="BIC",] %>%
left_join(BIC_panel_forward, by = c("set","model")) %>%
pivot_longer(cols = starts_with("X"), names_to = "var")) %>%
mutate(var = parse_number(var),
value = factor(value)) %>%
left_join(joined$policy_dic %>%
filter(policy_code %in% policy_raw) %>%
mutate(var = 1:n()),
by = "var") %>%
left_join(res_tab[,1:3] %>%
mutate(set = 1:n() %>% as.character),
by = "set") %>%
mutate(lags = case_when(optim_lag == -1 ~ 1,
optim_lag == -5 ~ 2,
optim_lag == -10 ~ 3),
scen = factor(scen,
levels = c("Low","Mid","High"),
labels = c("Any Efforts",
"Multi-level Efforts",
"Max. Efforts")),
max_date = factor(max_date,
levels = c("2020-04-13",
"2020-06-22"),
labels = c("Truncated",
"Full"))) -> var_select_res_forward
# plotting
var_select_res_forward %>%
filter(scen != "Multi-level Efforts") %>%
left_join(p_val_forward %>%
separate(var,
sep = "\\.",
into = c("policy_code", "component")),
by = c("policy_code",
"set",
"criterion")) %>%
#filter(max_date == "2020-04-13") %>%
ggplot(.) +
geom_rect(aes(xmin = lags-0.5,
ymin = var-0.5,
xmax = lags+0.5,
ymax = var+0.5,
fill = value)) +
geom_text(aes(x = lags,
y = var,
label = p_lab)) +
# facet_grid(scen ~ max_date + criterion) +
ggh4x::facet_nested(scen ~ max_date + criterion) +
geom_point(aes(x = 1, y = 5, color = cat), alpha = 0) +
scale_color_manual(values = c('#a6cee3',
'#1f78b4',
'#b2df8a',
'#33a02c')) +
geom_segment(data = data.frame(y = rep(0.5,4),
yend = rep(13.5,4),
x = seq(0.5,3.5,1),
xend = seq(0.5,3.5,1)),
aes(x = x, xend = xend, y = y, yend = yend)) +
geom_segment(data = data.frame(y = seq(0.5,13.5,1),
yend = seq(0.5,13.5,1),
x = rep(0.5,14),
xend = rep(3.5,14)),
aes(x = x, xend = xend, y = y, yend = yend)) +
theme_cowplot() +
# theme_bw() +
scale_y_continuous(breaks = seq(1,13,1),
labels = joined$policy_dic %>%
filter(policy_code %in% policy_raw) %>%
pull(lab),
trans = "reverse") +
scale_x_continuous(breaks = c(1,2,3),
labels = c("-1", "-5", "-10")) +
theme(axis.text = element_text(size = 20),
axis.title = element_text(size = 20),
legend.text = element_text(size = 20),
legend.title = element_text(size = 20),
legend.position = "bottom",
strip.background = element_rect(fill = "white",
color = "black"),
axis.text.y = element_text(color = c(rep('#a6cee3',7),
'#b2df8a',
rep('#1f78b4',2),
rep('#33a02c',3))),
legend.box = "vertical",
legend.box.just = "left",
legend.margin=margin(),
text = element_text(family = "Times New Roman"),
strip.text.y = element_text(face = "italic"),
strip.text = element_text(size = 20)) +
scale_fill_manual(values = c("snow2","darkgrey"),
labels = c("Variable Excluded","Variable Chosen")) +
labs(x = "Temporal Lags",
y = "",
fill = "Optimal Model",
color = "Intervention Category") +
guides(color = guide_legend(override.aes = list(alpha = 1,
size = 3),
nrow = 2)) -> p4_forward
here("figs/R2R", "fig4_forward.png") %>%
ggsave(width = 16, height = 10, plot = p4_forward)