Plot cumulative size of NF Data Portal over time

plot_portal_size.R

library(synapser)
library(tidyverse)
synLogin()

metadata <- synTableQuery("select createdOn, dataFileSizeBytes from syn16858331 where type = 'file'")$filepath %>% 
  read_csv() %>% filter(!is.na(dataFileSizeBytes)) %>% 
  mutate(cumulative_size=cumsum(dataFileSizeBytes)/1e12) %>% 
  mutate(cumulative_count=row_number()) %>% 
  mutate(date = createdOn/1000) %>% 
  mutate(date = lubridate::as_date(lubridate::as_datetime(date, origin = lubridate::origin)))

sf <- 1500

ggplot(metadata) + 
  geom_line(aes(x=date, y=cumulative_size), color = "red") +
  geom_line(aes(x=date, y=cumulative_count/sf), color = "blue") +
  scale_x_date(date_breaks = "1 year", date_labels =  "%Y") +
  scale_y_continuous(sec.axis = sec_axis(trans = ~.*sf, name = "# of files", labels = scales::comma)) +
  labs(x = "Year", y = "Cumulative Size (TB)") +
  theme_minimal() +
  theme(
      axis.title.y.left=element_text(color="red"),
      axis.text.y.left=element_text(color="red"),
      axis.title.y.right=element_text(color="blue"),
      axis.text.y.right=element_text(color="blue")
    )