Held: Tuesday, 28 January 2014

Back to Outline 05 - Drawings as Values. On to Outline 07 - Writing Your Own Procedures, Continued.

Summary

We begin to consider how you can write your own procedures and why you might do so.

Related Pages

• Reading: Drawings as Values
• Reading: Writing Your Own Procedures
• Lab: Drawings as Values
• Lab: Writing Your Own Procedures
• EBoard (Source) (HTML) (PDF)

Overview

• Why define your own procedures?
• How to define your own procedures.

Administrivia

• Note that if you use the "Source" version of an EBoard, you should see the stuff within a minute or two of me updating it.
• Since yesterday's lab went more smoothly than I expected, we're combining the drawings lab and procedure lab days. We'll continue with procedures tomorrow.
• Drawings lab writeup (due Friday): Exercise 4, parts a, c, d, f, and g
  • Email subject: CSC 151 Writeup 3: Drawings as Values (YOUR NAME(s) HERE)
• Procedures lab writeup (also due Friday): Exercises 3 and 6
  • Email subject: CSC 151 Writeup 4: Procedures (YOUR NAME(s) HERE)
• Extra credit:
  • Thursday at 4:15: Spencer Liberto, Lea Marolt Sonnenschein, and Daniel Torres on Ushahidi and more.
  • CS Table Friday at Noon: The ACM Code of Ethics.
  • Convo Feb. 5. (I'll give my "Why go to convo" lecture closer to the date.)
  • Others?

User Defined Procedures

• It's clear that programmers often want to (and need to) define their own procedures.
• By "their own procedures", we mean collections of Scheme commands that are parameterized and referred to be a single name, just like the built-in procedures, such as square and +, or the DrFu procedures, such as drawing-hshift.
• Procedures take inputs (which we call parameters) and may produce a result.
• Some procedures modify their parameters:
  • An "open jar" procedure changes the state of the jar
  • image-fill! adds virtual paint to an image.
• Some procedures create new values, without modifying their parameters:
• User-defined procedures can add clarity to a program.
  • Rather than looking at how code does something, the user of a procedure can focus on what the code does.
  • A reader of the program is much more likely to understand a procedure call than the body of the procedures.
• Programmers can avoid repetitive (and, therefore, error prone) code.
  • Rather than retyping the same code again and again, just changing a few values, a programmer can give a name to the same code.

• How do you define your own procedures? Using the following template:

  (define your-procedure (lambda (param₁ ... param_n) expression₁ ... expression_m))

• For example,

  (define square (lambda (val) (* val val)))

• You can (and should) document your procedures so that others can understand what they are supposed to do. We'll come back to this issue soon.

• When the body of a procedure has multiple expressions (as in the template), only the value of the last expression is returned.
• When you call a procedure, it's like you do a define for each parameter as the corresponding sent value.

• So (square 5) is a lot like

  (define val 5) (* 5 5)

Lab on User-Defined Procedures

• Begin the lab
• Be prepared to reflect (e.g., to describe the most important or most confusing thing you dealt with today). (And no, you can't say "Sam is the most confusing thing I dealt with today.")