Practice Activity Week 3

Accessing the Practice Activity

Download the template Practice Activity Quarto file here: pa-3.qmd

Important

Be sure to save the file inside your Week 3 folder of your STAT 431 (or 541) folder!

library(httr)
library(tidyjson)
library(jsonlite)
library(tidyverse)

Open-Notify API

The Open-Notify APIs provides information on the International Space Station. Specifically, the API provides (1) the names of the astronauts currently aboard the ISS, (2) the current location of the ISS, and (3) overhead pass predictions for the ISS.

For this activity, we are going to find the predicted time the ISS will pass by 10 major cities across the world and sort the pass times to see the path of the ISS.

City	Country	Latitude	Longitude
New York	USA	40.7128	-74.0060
Mexico City	Mexico	19.4326	-99.1332
São Paulo	Brazil	-23.5505	-46.6333
London	United Kingdom	51.5074	-0.1278
Cairo	Egypt	30.0444	31.2357
Nairobi	Kenya	-1.2921	36.8219
Mumbai	India	19.0760	72.8777
Bangkok	Thailand	13.7563	100.5018
Tokyo	Japan	35.6762	139.6503
Sydney	Australia	-33.8688	151.2093

Getting the Pass Times for One City

Inspecting the satellite pass API homepage, we can obtain the JSON data with the pass information for a given city using an address like the following: https://api.g7vrd.co.uk/v1/satellite-passes/25544/51.45/-2.5833.json?hours=18

In this web address,

• 25544 stands for the NORAD Catalog Number for the ISS. So, this number will never change for different cities.
• 51.45 is the latitude for the given location
• -2.5833 is the longitude for the given location
• ?hours=18 tells the API how many hours you want to consider for the passes

Let’s give this a go!

1. Using the base_url, lat and long objects, obtain the ISS pass times for New York City. Save this object as nyc. Hint 1: The glue() function from the glue package or the str_c() function from stringr will be helpful here! Hint 2: What function do you use to get information from an API?

base_url <- "https://api.g7vrd.co.uk/v1/satellite-passes/25544/"

lat  <- 40.7128
long <- -74.0060

2. Now that we have the JSON file for NYC, let’s parse it into a tidy dataframe. Using functions from jsonlite or tidyjson convert the JSON file to a dataframe. Hint: You might find it helpful to inspect the raw JSON file here.

3. The start column for the passtimes is what we are interested in, but currently this is formatted as a string (e.g., "2026-02-26T05:23:21.257Z"). Parse this column into a data-time object using the ymd_hms() function from lubridate. Hint: The pass times for the ISS are reported in UTC (Coordinated Universal Time).

4. Our final step is to isolate the pass time for NYC. Use the pull() function to extract the parsed data-time column you made in the last question. Can you collapse all the steps into one pipeline?

Getting the Pass Times for Every City

Now that we have worked out each step using one city, the next step is to scale this process to multiple cities. We’ve created a tibble with ten major cities across the world. Your goal is to acquire the pass time for each city.

cities <- tibble(
  city = c(
    "New York", "Mexico City", "Sao Paulo", "London",
    "Cairo", "Nairobi", "Mumbai", "Bangkok",
    "Tokyo", "Sydney"
  ),
  country = c(
    "USA", "Mexico", "Brazil", "United Kingdom",
    "Egypt", "Kenya", "India", "Thailand",
    "Japan", "Australia"
  ),
  latitude = c(
    40.7128, 19.4326, -23.5505, 51.5074,
    30.0444, -1.2921, 19.0760, 13.7563,
    35.6762, -33.8688
  ),
  longitude = c(
    -74.0060, -99.1332, -46.6333, -0.1278,
    31.2357, 36.8219, 72.8777, 100.5018,
    139.6503, 151.2093
  )
)

5. Our first step is to write a function that gets the pass time for any latitude and longitude coordinates. Using your code from Questions 1 though 4, create a get_pass_time() function which outputs the pass time for a given location. Hint 1: Your function should take two arguments: latitude and longitude. Hint 2: Think about how many objects you need to save!

6. Using map2() iterate your function over the latitude and longitude columns of the cities dataframe. Hint: There are three arguments to map2(), .x, .y, and .f. You will need to specify each!

Crap! It seems we got an error. Let’s think through what might be happening.

7. How does your function handle cities without pass times? What can you add to the body of your function to handle this special case?

8. Alright, let’s try this again! Using map2() iterate your function over the latitude and longitude columns of the cities dataframe.

9. Your map() should have been successful, but you should notice something else happening. There are cities with multiple pass times! What can you add to the body of your function to grab the first pass time for cities with multiple pass times?

10. Alright, let’s try this one last time! Using map2() iterate your function over the latitude and longitude columns of the cities dataframe. Save this object as city_passes so you don’t get an error about exceeding 10 requests per minute.

city_passes <- 

11. Add each city’s pass times to the original cities dataframe. Arrange the table based on the cities where the ISS will pass next.

Hint: The city_passes object is a list which you will need to unpack. The unnest() will be helpful here!

Rendering Your HTML

As you are likely aware, the API maxes out at 10 queries per minute. So, it will be impossible to render your HTML document without modifications (since you query the API 30 times!).

To render your HTML file, set #| eval: false for the code chunks associated with Question 6 and Question 8.