Get lists of all possible race/ethnicity categories and gender categories.
all.reth <- levels(nhanes_mec_svy$variables[, ridreth1])
all.gendr <- levels(nhanes_mec_svy$variables[, riagendr])Get smoothed density estimates for age in months, by cycle for each gender and race/ethnicity, using the survey::svysmooth function.
grct <- 1
smth.year <- vector(mode="list", length=10)
for (gendr in all.gendr){
for (r in all.reth){
grsub <- subset(nhanes_mec_svy, riagendr==gendr & ridreth1==r & !is.na(ridexagm))
smth.list <- vector(mode="list", length=3)
ct <- 1
for (y in grsub$variables[, levels(sddsrvyr)]){ #by cycle
ysub <- subset(grsub, sddsrvyr==y)
tmp <- svysmooth(~ridexagm,
ysub,
ngrid=960)$ridexagm
smth.list[[ct]] <- c(tmp, list(year=rep(y, length(tmp$x))))
ct <- ct + 1
}
tmp.year <- rbindlist(lapply(smth.list, as.data.table))
tmp.year[, gender:=gendr]
tmp.year[, reth:=r]
smth.year[[grct]] <- tmp.year
grct <- grct+1
}
}
#bind with smoothed data for all cycles combined
smth.all <- rbind(rbindlist(smth.year),
age_dist_smooth[, smth[[1]], by=.(gender, reth)])
smth.all[, gender:=factor(gender, levels=all.gendr)]
smth.all[, reth:=factor(reth, levels=all.reth)]Plot smoothed density estimates of age, by cycle and lumped.
#Plot
colvect <- c(RColorBrewer::brewer.pal(3, "Set2"), "black")
names(colvect) <- c('NHANES 2007-2008',
'NHANES 2009-2010',
'NHANES 2011-2012',
'Overall')
p <- ggplot(data=smth.all) +
geom_line(aes(x=x, y=y, color=year)) +
scale_color_manual(values=colvect) +
labs(x='Age, months',
y='Smoothed density estimate') +
facet_grid(reth~gender)+
theme_bw() +
theme(strip.background=element_blank())
print(p)
ggsave(filename="pdf_figures/age_smooth.pdf",
plot=p,
width=8.5,
height=11)