Lab 7: Data Types. Types of Data.

Data? Types?

We’ve already seen a bunch of data types that have been built into Pyret, such as String, Number, Table, and Boolean. We’ve even used our own custom data types, like TimeData and Posn! By creating custom data types, we can describe a piece of data with many components.

Today, we’ll do another design problem that draws on datatypes, tables, and lists.

Part 1: Shopping Spree

For this case, we’re asking you to go undercover and join a rag-tag band of rascals. In order to blend in, you’ll need to buy a disguise.

You head to your favorite clothes shoppe, which provides its catalog as an unsophisticated spreadsheet. You start sending in orders by carrier pigeon (they weren’t extinct yet), but the shoppe’s pigeons take a long time to come and have strange totals. You persuade the owners of the shoppe to give you permission to take a closer look at the data organization being used under the hood.

The first thing you notice is that the catalog spreadsheet gets loaded into a table, but the dimensions are stored as strings rather than in a format that makes it easy to determine item shapes and volumes for shipping purposes. You set out to fix this problem.

1. Import the catalog google sheet into your program as a table by copying the following lines of code into the Pyret definitions window:

   include shared-gdrive("cs111-2019.arr", "1anLZcL_Qr11cdmD2KO8UomSGb6pfOHV8")
   include gdrive-sheets
   import lists as L
   import math as M
   
   ssid = "1pNe1SQXe1IGOgI3XyAU4ugk9Ko5T3gzebJZZK-LZ7MM"
   data-sheet = load-spreadsheet(ssid)
   item-table =
     load-table: item-id, item-description, price, weight, dimensions
       source: data-sheet.sheet-by-name("Sheet1", true)
     end
   

2. Create a Dimension datatype, and transform each String in the “dimensions” column of the loaded catalog table into a Dimension (use transform-column).

   NOTE: the original dimensions column is formatted as: "WIDTHxDEPTHxHEIGHT"

   You’ll want to use the function string-to-number to convert the original dimensions from Strings to Numbers. Since someone might try to convert a String that contains characters other than digits (ex: string-to-number("!@#")), string-to-number returns a special type (called an Option) that signals whether the conversion worked. Here’s how to use string-to-number to extract the Number for a String of only digits (note the .value on the end):

   >>> string-to-number("!@#")
   none
   >>> string-to-number("42")
   some(42)
   >>> string-to-number("42").value
   42
   

CHECKPOINT: Call over a TA once you reach this point.

Part 2: Online Ordering

Next, you get a peek at how the shoppe is storing its orders. Each day, one table is created. That table contains a “name” column, specifying the person who placed the order, and a “description” column, listing the item that the corresponding person bought. Orders are bundled together and shipped on horseback each evening, so one shipment goes to each person whose name appears in the table (for instance, in the table below, John won’t receive two separate orders). Here is an example:

Gift cards (a relatively new technology) don’t appear in the catalog because they are by word-of-mouth, but people can still include them in their orders.

However, being a methodical and careful sheriff, you are quite unsettled by this method of tracking orders. They know that the shoppe needs to be able to perform several tasks with its data:

• Compute the total cost of each person’s order.
• Compute the total volume (with respect to spatial dimensions) of each person’s order.
• Compute the maximum single dimension of each person’s order (determines box size for shipping).
• Compute their total sales per day.
• Compute how many of each item gets sold for purposes of inventory.

Look at the two tables (catalog and orders), and keep these tasks in mind as you work on the following problems:

1. Discuss with your partner the strengths and weaknesses of the current data organization. Write down your main concerns as a collection of brief bullet points.

2. Spend 5-10 minutes coming up with an alternative proposal for the datatypes, tables, and lists that you might use. Indicate the types of all of your columns, components, and list contents. How does your proposal address each of your concerns from the previous question?

3. Now, take a look at two of our concrete proposals, listed in this document. Which do you prefer and why?

CHECKPOINT: Call over a TA once you reach this point.

Part 3: The Tables Have Turned

“Well, well, well, how the turntables…”

To get more practice working with datatypes, we will now work with the second of our proposals.

1. Define the datatypes for each of Order, UserOrder, and ItemData as described in that proposal.

2. How specifically does this collection of datatypes address the issues that we identified with the original shoppe design?

3. Recreate the information in the above example orders table with your new datatypes (name it orders).

   NOTE: you should use all three of the datatypes you defined, in addition to the dimension datatype.

   NOTE: you do not need to create a new table using these datatypes. We want you to rewrite the data represented in the table with your datatypes. For example, to represent John’s orders, we can write something like

   john = user-order("John", [list: order(fuzzy-socks, 5), order(lamp, 1)])
   

4. Write a function any-oversized that takes a List<Order> and returns a Boolean indicating whether any single item in the order has a total linear dimension (length + width + height) of more than 40 inches (the same units as the original table).

5. Write a function more-socks that takes a List<Order> (that does not contain any gift cards) and returns a List<Order> that has the items from the original orders, except each item matching the description “Fuzzy Socks” has its count increased by 1.

6. Write a function order-cost that takes a List<UserOrder> and a customer name and returns the total price of the items in that person’s orders. Gift cards cost their amount plus a 50-cent processing fee. Physical items cost the price associated with the item (ignore tax, shipping costs, etc). Assume that the list contains the customer.

CHECKPOINT: Call over a TA once you reach this point.

Crooks caught!

As usual, you’ve solved the case in record time. The undercover operation was a success and all the crooks have been caught thanks to your quick coding.