Exercise 9: Transform a collection

Aim: map, select and reduce values while keeping intermediate results visible. Save ex09.fib.

Type and predict

(ns main (:use fib.core fib.seq fib.coll))
(defun main () -> i64
  (let ((numbers [1 2 3 4])
        (squares (mapv (fn (x: i64) (* x x)) numbers))
        (large (filterv (fn (x: i64) (> x 4)) squares))
        (total (reduce + 0 large)))
    (do
      (println numbers)
      (println squares)
      (println large)
      (println total)
      0)))

fn creates an unnamed function. mapv returns a vector with a result for each input; filterv returns a vector of inputs satisfying the predicate. reduce combines values, starting with the explicit initial value 0. The v matters: map returns a sequence, which is a different type from a vector.

Check your prediction

[1 2 3 4]
[1 4 9 16]
[9 16]
25
0

Change it

1. Change the threshold to 100. Predict the empty selection and its total.
2. Filter the original numbers before squaring. Explain how “square then select values above four” differs from “select inputs above four then square.”
3. Replace the anonymous squaring function with a named square function. Require unchanged output.

Break and repair

Change the reduction's initial value to 10. It runs but produces the wrong answer for a sum. Test both a nonempty and an empty selection, then restore zero. This is a behaviour mistake; a compiler cannot infer the sum you intended.

Checkpoint

Give the type and element count of each intermediate value. Explain why an empty reduction needs an initial value. Reference: collections.

← Loops · Course · Alternatives →