/
04_scale_spectrum_output_age_sex_specific.R
executable file
·208 lines (173 loc) · 8.91 KB
/
04_scale_spectrum_output_age_sex_specific.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
rm(list=ls())
gc()
jpath <- ifelse(Sys.info()[1]=="Windows", "FILEPATH", "FILEPATH")
list.of.packages <- c("data.table","ggplot2","parallel","gtools","haven")
new.packages <- list.of.packages[!(list.of.packages %in% installed.packages()[,"Package"])]
if(length(new.packages)) install.packages(new.packages)
lapply(list.of.packages, require, character.only = TRUE)
jpath <- ifelse(Sys.info()[1]=="Windows", "FILEPATH", "FILEPATH")
hpath <- ifelse(Sys.info()[1]=="Windows", "FILEPATH", "FILEPATH")
code.dir <- ifelse(Sys.info()[1]=="Windows", "FILEPATH", "/homes/")
cores <- 5
source(paste0(jpath,"FILEPATH/multi_plot.R"))
source(paste0(jpath, "FILEPATH/get_population.R"))
### Arguments
if(!is.na(commandArgs()[3])) {
c.fbd_version <- commandArgs()[3]
c.iso <- commandArgs()[4]
c.scenario <- commandArgs()[5]
} else {
c.fbd_version <- "20170727"
c.iso <- "ROU"
c.scenario <- "reference"
}
print(commandArgs())
print(c.fbd_version)
print(c.iso)
print(c.scenario)
c.args <- fread(paste0(code.dir,"FILEPATH/run_versions.csv"))
c.args <- c.args[fbd_version==c.fbd_version]
c.gbd_version <- c.args[["gbd_version"]]
extension.year <- c.args[["extension_year"]]
c.draws <- c.args[["draws"]]
gbd.aggs.version <- c.args[["gbd_aggs_version"]]
gbd_version <- "FILEPATH"
fbd_version <- "FILEPATH"
#load gbd and fbd data
gbd <- fread(paste0(gbd_version,c.iso,".csv"))[run_num %in% 1:c.draws]
fbd <- fread(paste0(fbd_version,c.iso,"_ART_data.csv"))[run_num %in% 1:c.draws]
loc_id <- unique(gbd[,location_id])
# collapse older gbd
gbd[age_group_id > 20, age_group_id := 21]
gbd <- gbd[, lapply(.SD, sum), by = .(year_id, sex_id, age_group_id, run_num, location_id)]
#convert fbd variables to match gbd
#sex
fbd[sex=="male",sex_id:=1]
fbd[sex=="female",sex_id:=2]
#age
fbd[,age_group_id:=(age/5)+5]
temp4 <- fbd[age_group_id==5]
temp3 <- fbd[age_group_id==5]
temp2 <- fbd[age_group_id==5]
temp4[,age_group_id:=4]
temp3[,age_group_id:=3]
temp2[,age_group_id:=2]
fbd <- rbind(fbd,temp4,temp3,temp2)
#pops
fbd[,pop:=pop_neg+pop_lt200+pop_200to350+pop_gt350+pop_art]
fbd[,pop_hiv:=pop - pop_neg]
setnames(gbd, c("population", "year_id"), c("pop", "year"))
#make rates
gbd[,g_mort_rate:=hiv_deaths/pop]
gbd[,g_inc_rate:=scaled_new_hiv/pop]
gbd[,g_prev_rate:=pop_hiv/pop]
gbd[,g_art_rate:=ifelse(pop_hiv == 0, 0, pop_art/pop_hiv)]
fbd[,f_mort_rate:=hiv_deaths/pop]
fbd[,f_inc_rate:=new_hiv/pop]
fbd[,f_prev_rate:=pop_hiv/pop]
fbd[,f_art_rate:=ifelse(pop_hiv == 0, 0, pop_art/pop_hiv)]
#fill in missing values
fbd <- fbd[order(age,sex,run_num,year)]
fbd[,temp_f_art_rate:=((shift(f_art_rate,n=1L,type="lag") +shift(f_art_rate,n=1L,type="lead")) / 2) ]
fbd[,temp_f_inc_rate:=((shift(f_inc_rate,n=1L,type="lag") +shift(f_inc_rate,n=1L,type="lead")) / 2) ]
fbd[,temp_f_prev_rate:=((shift(f_prev_rate,n=1L,type="lag") +shift(f_prev_rate,n=1L,type="lead")) / 2) ]
fbd[,temp_f_mort_rate:=((shift(f_mort_rate,n=1L,type="lag") +shift(f_mort_rate,n=1L,type="lead")) / 2) ]
fbd[,temp2_f_art_rate:=((shift(f_art_rate,n=2L,type="lag") +shift(f_art_rate,n=2L,type="lead")) / 2) ]
fbd[,temp2_f_inc_rate:=((shift(f_inc_rate,n=2L,type="lag") +shift(f_inc_rate,n=2L,type="lead")) / 2) ]
fbd[,temp2_f_prev_rate:=((shift(f_prev_rate,n=2L,type="lag") +shift(f_prev_rate,n=2L,type="lead")) / 2) ]
fbd[,temp2_f_mort_rate:=((shift(f_mort_rate,n=2L,type="lag") +shift(f_mort_rate,n=2L,type="lead")) / 2) ]
fbd[is.na(f_inc_rate),f_inc_rate:=temp_f_inc_rate]
fbd[is.na(f_inc_rate),f_inc_rate:=temp2_f_inc_rate]
fbd[is.na(f_mort_rate),f_mort_rate:=temp_f_mort_rate]
fbd[is.na(f_mort_rate),f_mort_rate:=temp2_f_mort_rate]
fbd[is.na(f_prev_rate),f_prev_rate:=temp_f_prev_rate]
fbd[is.na(f_prev_rate),f_prev_rate:=temp2_f_prev_rate]
fbd[is.na(f_art_rate),f_art_rate:=temp_f_art_rate]
fbd[is.na(f_art_rate),f_art_rate:=temp2_f_art_rate]
gbd[g_prev_rate==0,g_art_rate:=0]
fbd[f_prev_rate==0,f_art_rate:=0]
#subset variables
gbd <- gbd[,.SD,.SDcols=c("pop", "year","age_group_id","sex_id","run_num","g_mort_rate","g_inc_rate","g_prev_rate","g_art_rate")]
gbd <- melt.data.table(gbd, id.vars = c("year", "age_group_id", "sex_id", "run_num", "pop"))
fbd <- fbd[,.SD,.SDcols=c("pop","year","age_group_id","sex_id","run_num","f_mort_rate","f_inc_rate","f_prev_rate","f_art_rate")]
fbd <- melt.data.table(fbd, id.vars = c("year", "age_group_id", "sex_id", "run_num", "pop"))
### Calculate draw-specific scalars with age-standardized rates
# Read in GBD 16 standard age weights
weights <- fread(paste0(jpath,'FILEPATH.csv'))
## FBD
# Calcualte both-sexes rate in extension year
fbd.both <- fbd[year_id == extension.year, .(value = weighted.mean(x = value, w = pop)), by = .(age_group_id, run_num, variable)]
fbd.ext <- merge(fbd.both, weights, by = c("age_group_id"))
prev <- fbd.ext[variable == "f_prev_rate"]
prev[,prev_count:=age_group_weight_value*value]
prev <- prev[,.SD,.SDcols=c("age_group_id","run_num","prev_count")]
#merge on prev counts as denom for ART
fbd.ext <- data.table(merge(fbd.ext,prev,by=c("age_group_id","run_num"),all.x=T))
fbd.ext[variable=="f_art_rate",age_group_weight_value:=prev_count]
#calculate national all-age both sex agg draws
fbd.aggs <- fbd.ext[,.(value=weighted.mean(x=value,w=age_group_weight_value)),by=.(run_num,variable)]
fbd.aggs[is.na(value), value := 0]
## GBD
gbd.both <- gbd[year == extension.year, .(value = weighted.mean(x = value, w = pop)), by = .(age_group_id, run_num, variable)]
gbd.ext <- merge(gbd.both, weights, by = c("age_group_id"))
prev <- gbd.ext[variable == "g_prev_rate"]
prev[,prev_count:=age_group_weight_value*value]
prev <- prev[,.SD,.SDcols=c("age_group_id","run_num","prev_count")]
#merge on prev counts as denom for ART
gbd.ext <- data.table(merge(gbd.ext,prev,by=c("age_group_id","run_num"),all.x=T))
gbd.ext[variable=="g_art_rate",age_group_weight_value:=prev_count]
#calculate national all-age both sex agg draws
gbd.aggs <- gbd.ext[,.(value=weighted.mean(x=value,w=age_group_weight_value)),by=.(run_num,variable)]
gbd.aggs[is.na(value), value := 0]
## Combine
scale.dt <- rbind(fbd.aggs, gbd.aggs, use.names = T)
scale.dt <- dcast.data.table(scale.dt, run_num ~ variable, value.var = "value")
scale.dt[f_art_rate != 0,f_art_rate := f_art_rate]
scale.dt[g_art_rate!= 0,g_art_rate := g_art_rate]
# calculate scalars
scale.dt[,scale_mort_rate:= ifelse(f_mort_rate == 0, 0, g_mort_rate / f_mort_rate) ]
scale.dt[,scale_inc_rate:= ifelse(f_inc_rate == 0, 0, g_inc_rate / f_inc_rate) ]
scale.dt[,scale_prev_rate:= ifelse(f_prev_rate == 0, 0, g_prev_rate / f_prev_rate) ]
scale.dt[,scale_art_rate:= ifelse(f_art_rate == 0, 0, g_art_rate / f_art_rate) ]
scale.dt <- scale.dt[,.SD,.SDcols=c("run_num","scale_mort_rate","scale_inc_rate","scale_prev_rate","scale_art_rate")]
# merge on scalars to combined dataset and apply them in the future
setnames(gbd, "year", "year_id")
fgbd <- rbind(fbd, gbd, use.names = T)
fgbd[, pop := NULL]
fgbd <- dcast.data.table(fgbd, year_id + age_group_id + sex_id + run_num ~variable, value.var = "value")
fgbd <- merge(fgbd, overlap, by = c("run_num"))
fgbd[,f_mort_rate:=f_mort_rate * scale_mort_rate]
fgbd[,f_inc_rate:=f_inc_rate * scale_inc_rate]
fgbd[,f_prev_rate:=f_prev_rate * scale_prev_rate]
fgbd[f_art_rate != 0,f_art_rate:=f_art_rate * scale_art_rate]
fgbd[f_art_rate >1, f_art_rate := 1]
#take gbd in past, fbd in future
fgbd[year_id<= extension.year,mort_rate:=g_mort_rate]
fgbd[year_id>extension.year,mort_rate:=f_mort_rate]
fgbd[year_id<= extension.year,art_rate:=g_art_rate]
fgbd[year_id>extension.year,art_rate:=f_art_rate]
fgbd[year_id<= extension.year,prev_rate:=g_prev_rate]
fgbd[year_id>extension.year,prev_rate:=f_prev_rate]
fgbd[year_id<= extension.year,inc_rate:=g_inc_rate]
fgbd[year_id>extension.year,inc_rate:=f_inc_rate]
#reshape long on var, wide on draws for disk size
fgbd <- fgbd[,.SD,.SDcols=c("year_id","run_num","mort_rate","inc_rate","prev_rate","art_rate", "age_group_id", "sex_id")]
fgbd <- melt.data.table(fgbd,id.vars=c("run_num","year_id", "sex_id", "age_group_id"))
#fill tiny number of missing values with 0
missing_percent <- nrow(fgbd[is.na(value)]) / nrow(fgbd)
if (missing_percent < .05) {
fgbd[is.na(value),value:=1e-10]
}
fgbd[variable=="inc_rate",variable:="Incidence"]
fgbd[variable=="prev_rate",variable:="Prevalence"]
fgbd[variable=="mort_rate",variable:="Mortality"]
fgbd[variable=="art_rate",variable:="ART Coverage"]
fgbd[,ihme_loc_id := c.iso]
fgbd[,location_id := loc_id]
###make folders, save
dir.create(paste0(jpath,"FILEPATH"),showWarnings = F,recursive = T)
write.csv(fgbd,paste0(jpath,"FILEPATH",c.iso,".csv"))
#calculate national all-age both sex agg sumarries
summary <- fgbd[,.(mean=mean(value),upper=quantile(value,probs=c(.975)),lower=quantile(value,probs=c(.025))),by=.(year_id,location_id,ihme_loc_id,variable, age_group_id, sex_id)]
dir.create(paste0(jpath,"FILEPATH"),showWarnings = F,recursive = T)
write.csv(summary,paste0(jpath,"FILEPATH/",c.iso,".csv"))