CSC 151.03, Class 16: Preconditions, revisited

Overview

• Preliminaries
  • Notes and news
  • Upcoming work
  • Extra credit
  • Friday PSA
  • Questions
• Quiz
• Lab

News / Etc.

• New partners.
• Please make sure to return your computer cards to the jar.
• I was sorry to see so few of you at the Wilson Okello convo. It was one of the better convocations I’ve seen.

Upcoming Work

• Writeup for class 15 due TONIGHT at 10:30 p.m.
  • Exercise 5
  • To: csc151-03-grader@grinnell.edu
  • Subject: CSC 151.03 Writeup 15 (YOUR NAMES)
• Writeup for class 16 due Monday at 10:30 p.m.
  • Exercise 2
  • To: csc151-03-grader@grinnell.edu
  • Subject: CSC 151.03 Writeup 16 (YOUR NAMES)
• Assignment 5 due Tuesday at 10:30 p.m.
• Reading for Mondays’s class: Recursion basics

Extra credit (Academic/Artistic)

• Any of the many Grinnell Prize events.
• CS Table, Tuesday: Tapia.

Extra credit (Peer)

• ISO Fundraiser for Rohingya crisi in Myanmar, Friday, 5:30 p.m. JRC 209. (You need not donate to attend.) (Henna tatoos.)

Other good things

• Women’s Tennis vs. Cornell, Noon, Sunday (high-school courts)
• Volleyball vs. Knox Tuesday at 7:00 p.m.
• Volleyball vs. Beloit, next Friday at 7:00 p.m.
• Men’s Soccer Wednesday at 4:30 vs. Cornell

Friday PSA

• Think about your choices, in advance.
• Self gov (individually and collectively)
• Consent is absolutely positively necessary. But you should strive for more than consent.

Questions

What is the difference between a Boolean and a predicate?

Booleans are basic values, like integers, reals, and strings. The two Boolean values are #t and #f. The design of Scheme is such that anything that’s not a Boolean is treated as “truish” and equivalent to #t in conditioanls.

Predicates are procedures that return Boolean values.

We put question marks after both predicates and Boolean values to highlight that they are or return true/false.

Can you help with problem 1 on the homework?

Sure. Let’s look at what the problem says.

(define explore-lengths
  (lambda (fname)
    (let* ([words (file->words fname)]
           [num-words (length words)]
           ; Given a (length frequency) pair, scale the
           ; frequency by the number of words.
           [scale-frequency (lambda (oh)
                              (list "hi" (/ 2 num-words)))])
      (map scale-frequency
           (tally-all (map (section string-ref <> 0) words))))))

I’m not sure what to put as the postconditions for explore-lengths.

Yeah, that’s a hard one. You can’t be as precise, but say something about the number of bars you see and such.

Kind of like the purpose, but perhaps more detail. Here, the purpose is “make a histogram of word frequencies” and the postcondition probably explains what each of the two potentially puzzling terms mean (“histogram” “frequency”)

For problem 1, what should I do about word lengths that aren’t included? I feel like it distorts the results.

Ideally, you’d come up with a clever solution. But it’s a rare enough case that I do not require you to come up with a solution.

Suppose I have data for a stacked histogram. How do I figure out the

number of entries in the legend?

You have ‘((“a” (1 2 3)) (“b” ()) (“c” (6 1 3 2)) (“d” (5 1 4)))

I can use (map cadr data) to extract the lists

And so on and so forth

> (map cadr data)
'((1 2 3) () (6 1 3 2) (5 1 4))
> (map length (map cadr data))
'(3 0 4 3)
> (reduce max (map length (map cadr data)))
4
> (define len (reduce max (map length (map cadr data))))

We’ll suppose that I have a predefined list of legends and colors. I can use (take legends len) and (take colors len)

Can you extend that example?

Sure

> (define all-labels (list "twas" "brillig" "and" "the" "slithy" "toves" "did" "gyre"))
> (define all-colors (list "red" "orange" "yellow" "green" "blue" "indigo" "violet" "black" "white" "gray" "purple" "pink"))
> (define data '(("a" (1 20 8)) ("b" (10 10)) ("c" (3 4 5 6)) ("d" (2))))
> (define len (reduce max (map length (map cadr data))))
> (plot (stacked-histogram data
                           #:colors (take colors len)
                           #:labels (take legends len)))
> (define data '(("a" (1 8)) ("x" ()) ("b" (5 5 5 5 5 5)) ("c" (3 4 5 6)) ("d" (2))))
> (define len (reduce max (map length (map cadr data))))
> (plot (stacked-histogram data
                           #:colors (take colors len)
                           #:labels (take legends len)))

Quiz

Lab

How do I define city? : (define city? (lambda (val) (and (list? val) ... (string? (list-ref val 0)) (real? (list-ref val 1)) ...)))

Are there other inputs I should try?

(city? (list "50112" 52 -90))

Do I have to verify that the zip code is five digits?

No.

Debrief