Faceted_Graphs_of_Age.R

projected_episodes_file <- 'scc-episodes-2019-08-13-rewind-1yr-train-3yr-project-5yr-runs-100-seed-42-20201203-no-reject-sampling.csv'
projected_episodes <- read.csv(projected_episodes_file)

periods <- projected_episodes %>%
  group_by(Simulation, ID) %>%
  slice(1) %>%
  ungroup

periods$Period.End <- ymd(periods$Period.End)
i=0

chart_data <- periods %>%
  mutate(sample_label = year(Period.End - months(7) - days(12)) - 2019) %>%
  filter(sample_label %in% -5:4) %>%
  # mutate(sample_label = if_else(sample_label < 0, "Historic", "Projected")) %>%
  mutate(sample_label = factor(paste(sample_label), levels = as.character(-5:4))) %>%
  # mutate(Period.Duration = Period.Duration + rnorm(1, sd = 7)) %>%
  group_by(sample_label, Admission.Age) %>%
  mutate(cdf = ecdf(Period.Duration)(Period.Duration)) %>%
  ungroup

output_file <- 'output-1.pdf'
pdf(output_file)
print(chart_data %>%
        filter(chart_data$Period.Duration<=2000)%>%
        ggplot(aes(Period.Duration, cdf, group = sample_label, colour = sample_label)) +
        facet_grid(sample_label~Admission.Age,scales='free')+
        theme(panel.grid.major = element_line(size = 0.1, linetype = "solid",colour = "darkgrey"))+
        theme(panel.grid.minor = element_line(size = 0.1, linetype = "solid",colour = "darkgrey"))+
        theme(axis.text.x = element_text(size = 3,angle=90,hjust = 1),
              axis.text.y =element_text(size = 3))+
        geom_vline(data = chart_data %>% group_by(sample_label,Admission.Age) %>% summarise(intercept = mean(Period.Duration)),
                   aes(xintercept=intercept,colour ="mean")) +
        geom_vline(data = chart_data %>% group_by(sample_label,Admission.Age) %>% summarise(intercept = median(Period.Duration)),
                   aes(xintercept=intercept,colour ="median")) +
        geom_line() +
        scale_color_manual(values = tableau_color_pal("Tableau 20")(20)) +
        labs(title = "CDF for joiners") +
        theme_mastodon)

dev.off()



new_table<-chart_data %>% group_by(sample_label,Admission.Age) %>% summarise(intercept = mean(Period.Duration))

max(new_table$intercept)