Assignment 3: Sunrise, Sunset

We love sunny days, and as winter begins to fade, the hours of daylight are beginning to grow. In this assignment, we’ll use our experience working with tables to process solar and lunar temporal information and to visualize how the hours of sunlight and moonlight vary over time.

include shared-gdrive("dcic-2021",
  "1wyQZj_L0qqV9Ekgr9au6RX2iqt2Ga8Ep")

And remember, for help with table functions, refer to this documentation, which goes with the dcic-2021 functions, rather than the official Pyret documentation.

Some kindly astronomers have shared a data set with us, which they’ve made available as a Google sheet.

To ensure the autograder works for your program, use the column names

year, month, day,
age-of-moon,
sun-rise, sun-culm, sun-set,
moon-rise, moon-culm, moon-set

We won’t use all of the columns in this spreadsheet. This is perfectly normal when working with real data sets!

test-data =
  table:
    year, month, day,
    sun-rise, sun-set,
    moon-rise, moon-set
    row: 2021, 9, 30, "06:30", "18:30", "00:00", "16:00"
    row: 2022, 3, 22, "07:00", "18:00", "23:00", "08:30"
  end

You should use your test-data table to write examples for each function in a where block. While the autograder won’t deduct points for missing tests, your instructor will! Doing so before you write the body of the function will ensure you understand how the function is supposed to work.

To compute the hours of sunlight, we’ll need to know the elapsed time between sunrise and sunset. It’s complex to do this with times expressed as hours and minutes, e.g., "06:30", so we’ll first write a function to convert a time to minutes.

fun mins-since-midnight(time :: String) -> Number:
  ...
end

Hints:

• For each function on the assignment, try writing examples in the where block before you write the body of the function. This ensures you understand how the function is supposed to work!
• Since the times are strings, you may want to review the documentation listing the functions Pyret provides for working with strings. How could you find just the hours in the string? Just the minutes? Some of the string functions you’ll see in the documentation produce a list. Although you may have seen lists in class, you don’t need to use lists to solve this problem!

• After you’ve identified the hours and minutes, you’ll need to use the function string-to-number. Despite its name, the output is not a number! It’s some number if the string can be converted and none if it can’t. For this assignment, the strings can always be converted to numbers, so you only need to worry about the some case. To get the number out, use .value:

  ››› string-to-number("06")
  some(6)
  ››› string-to-number("06").value
  6
  

Now that you can convert the sunrise and sunset times to minutes since midnight, it’s not too hard to compute how many minutes are in between.

fun elapsed-mins(start-time :: String, end-time :: String)
  -> Number:
  ...
end

And now we’re ready to add a new column to the table containing the hours of sunlight.

Remember: To do this, you’ll need to define a function that you’ll pass as an input to build-column, and it will need to convert the elapsed minutes to hours.

Now that we’ve computed the hours of sunlight, it seems like it should be simple to compute the hours of moonlight – just swap the column names "sun-rise" and "sun-set" for "moon-rise" and "moon-set".

This works for some of the rows in our table:

››› good-night = test-data.row-n(0)
››› elapsed-mins(good-night["moon-rise"], good-night["moon-set"])
960

But not for others:

››› bad-night = test-data.row-n(1)
››› elapsed-mins(bad-night["moon-rise"], bad-night["moon-set"])
-870

Why is this? The moon rose at 23:00 (i.e., 11 p.m.) on one day and set at 8:30 a.m. the next.

Let’s extend our code to handle this!

fun mins-before-midnight(time :: String) -> Number:
  ...
end

Now we can improve elapsed-mins. Let’s leave our original version (which works for sunlight) and define a new one:

fun elapsed-mins-improved(start-time :: String, end-time :: String)
  -> Number:
  ...
end

With elapsed-mins-improved, we’re ready to compute the hours of moonlight for each day in the table.

While we can look at the numbers we added to the table, it’s much nicer to get a graphic overview.

We’ll draw bar charts, but, unfortunately, none of the columns in the table give a good label for the bars. Since each bar is for one day, they should be labeled with the date.

bar-chart(celestial-data-to-plot, "date", "sunlight")

bar-chart(celestial-data-to-plot, "date", "moonlight")

The spreadsheet we loaded the table from in Part 1 actually has two sheets. The one we used, celestial is a cleaned version of the original data, which is on the sheet named tricky.

In that original data, some of the times are missing. This is because it’s possible for moon rise, moon culm., or moon set to be on a different day (e.g., the moon rose the day before or set the following day). We removed these troublesome rows for you in celestial.

As with Assignment 2, you are expected to follow good Pyret style, including writing docstrings and examples for each function. Review the Testing and Style Guidelines and ask questions if anything’s unclear!

1. Download your file (File → Download) and ensure it’s named asmt03.arr.
2. Upload your assignment on Gradescope.

Note: You can submit as many times as you want before the deadline. Only your latest submission will be graded.