time_in_space.R

library(httr)
library(jsonlite)
library(dplyr)
library(ggplot2)

# Download full database
astronaut_db_url <-
  'https://supercluster-iadb.s3.us-east-2.amazonaws.com/adb.json'
astronauts_db <-
  jsonlite::fromJSON(content(GET(astronaut_db_url), "text"))

# Save database to file
writeLines(jsonlite::toJSON(astronauts_db, auto_unbox = TRUE),
           'astro_db.json')

# Loading Data
astronauts_db <- jsonlite::fromJSON('astro_db.json')

# Extracting Time in Space
s_space_time <- astronauts_db$astronauts %>%
  filter(!is.na(totalMinutesInSpace)) %>%
  mutate(totalDaysInSpace = totalMinutesInSpace / (60 * 24)) %>%
  select(name, totalDaysInSpace) %>%
  arrange(totalDaysInSpace) %>%
  tibble()

#Filter out zero values
s_space_time_filtered <-
  s_space_time[s_space_time$totalDaysInSpace > 0,]
total_astronauts <- nrow(s_space_time_filtered)

# subtitle
summary_data <- summary(s_space_time_filtered$totalDaysInSpace)
percentage <- round(summary_data[3] / total_astronauts * 100, 2)
median_days <- round(summary_data[3], 1)
subtitle <- stringr::str_wrap(
  paste0(
    "Out of ",
    total_astronauts,
    " astronauts, ",
    total_astronauts / 2,
    " individuals, or 50 %,",
    " spent ",
    median_days,
    " days or less in space, which is the median duration of space missions."
  ),
  55
)

# Visualization
ggplot(s_space_time_filtered, aes(x = totalDaysInSpace)) +
  geom_histogram(fill = 'red',
                 color = "white",
                 binwidth = 0.1) +
  scale_x_continuous(
    trans = 'log10',
    breaks = c(0.01, 0.1, 1, 10, 25, 100, 1000),
    labels = c("1e-02", "1e-01", "1", "10", "25",  "100",  "1000")
  ) +
  scale_y_continuous(
    breaks = seq(0, 60, 10),
    expand = c(0, 0),
    limits = c(0, 65)
  ) +
  labs(
    x = 'Days in Space (log scale)',
    y = 'Number of Astronauts',
    title = 'Time Spent in Space',
    subtitle = subtitle,
    caption = "Author: ChatGPT\nSource: Astronaut Database - Supercluster.com"
  ) +
  theme(
    text = element_text(family = "PT Sans", size = 18),
    plot.title = element_text(color = 'white', size = 22),
    plot.subtitle = element_text(color = 'white', size = 16),
    plot.caption = element_text(color = 'white', size = 12),
    axis.text = element_text(color = 'white'),
    axis.title = element_text(color = 'white'),
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    panel.grid.major.y = element_line(color = 'white', size = 0.5),
    panel.background = element_rect(fill = 'black'),
    plot.background = element_rect(fill = 'black')
  )

ggsave(
  "time_spent_in_space.png",
  width = 6,
  height = 6,
  dpi = 300,
  scale = 1
)