A little room to wonder

Repeating Instructions & Loops

Trace repeated instructions by recording changing values and explicit boundaries rather than guessing from a short pattern.

What has to stay the same, and what has to change, each time a loop repeats?

A collection of coloured paperclips.
Repeated paperclips suggest a collection to visit one item at a time. The photograph is not the 3, 5, 2 trace or a specific computer program.Dietmar Rabich · Source · CC BY-SA 4.0

A loop combines a repeated action with a rule

A loop repeats an instruction block under a stated rule. It may visit every item in a finite sequence, repeat a fixed number of times or continue while a condition holds. Those rules answer different questions about how repetition begins and ends.

The repeated block can stay the same while its input or stored values change. Recording the current item, counter and result makes that change visible. A loop is therefore more than writing “do it again”: the algorithm must specify what again means and how the next repetition differs.

Worked example: trace the running total

Start an invented paper procedure with total = 0. Visit the numbers 3, 5 and 2 in that order, adding the current number to total each time. After the three visits, the totals are 3, 8 and 10. The final 10 represents the sum, while the intermediate values show how it was produced.

If total were reset to 0 inside the repeated block, the final result would be 2 instead. That is a different procedure even though the same three numbers were visited. The location of an initialisation step can matter as much as the addition rule.

Check the first and last repetition

Boundary mistakes often add or omit one repetition. In Python’s range notation, range(4) supplies 0, 1, 2, 3: four values, with 4 excluded. This article’s paper procedures do not depend on using Python, but that example shows why a boundary convention must be stated.

For a condition-controlled model, identify what can eventually make the condition false. An instruction to repeat while a count is below four will not finish if the count never changes. Tracing is a safe way to find that logic error without running unknown code or operating a device.

Test a different kind of input

The empty sequence is a useful case. A sum procedure starting from 0 can return 0 without visiting anything. A procedure seeking a largest item needs a different plan for an empty collection; copying the sum’s starting value would not solve that problem.

A convincing explanation names the task, starting state, repeated block and finishing rule. One successful trace supports the example, while additional cases test different assumptions. You can reason about all of these using written numbers rather than installing software.

Only if you feel like trying

Move the reset step on paper

  1. Trace the supplied 3, 5, 2 procedure with total initialised before the loop.
  2. Trace it again with the reset incorrectly inside the repeated block and explain the different result.
  3. Try an empty list and one-item list. State which result the sum rule should produce.

Optional. Use paper or talk through the supplied fictional example. No purchase, personal information, account, public sharing or practical experiment is needed.