Additional programming techniques: Random number generation

The idea

Random number generation produces values that are not predictable in advance within a chosen range, useful for games, simulations, sampling and testing. Most languages provide a library routine, so the complex generation algorithm need not be written from scratch. These routines are usually pseudorandom: values look random but are produced by a deterministic algorithm, so they are predictable if the seed and algorithm are known. Syntax and bounds are language-specific. In OCR Reference Language, `random(x, y)` returns an integer from x to y inclusive, so a die roll is `random(1, 6)`. In Python, after `import random`, `random.randint(1, 6)` returns an integer from 1 to 6 inclusive. Generated values can be stored in variables, tested with selection, used to index an array, or repeated inside loops.

What earns the marks

• States that random number generation produces values that are not predictable in advance within a defined range.
• Identifies how to call a library routine to generate a random number, giving the correct syntax and inclusive bounds for the language or pseudocode used.
• Applies random values to a context, for example simulating a dice roll with a random integer from 1 to 6, or selecting a random item from an array.
• Describes how generated values can be stored in variables, tested with selection statements or repeated inside loops.
• Recognises that random number generation is a built-in library routine in most programming languages, avoiding the need to write the generation algorithm from scratch.
• Uses OCR Reference Language correctly, writing `random(x, y)` for an integer from x to y inclusive, e.g. `random(1, 6)` for a die roll.
• Explains that most language routines are pseudorandom, so values are predictable if the seed and algorithm are known.

Common mistakes and how to put them right

• Assuming a random function can return a value outside its stated bounds.

  Put it right: State the bounds and note that both are inclusive, e.g. `random(1, 6)` returns 1 to 6.

• Writing `random(n)` in OCR Reference Language.

  Put it right: The routine takes two arguments, so write `random(1, 6)` for a die roll.

• Writing `randint(1, 6)` in Python without importing.

  Put it right: Write `import random` then `random.randint(1, 6)`, or `from random import randint`.

• Forgetting to store the generated random number in a variable before using it multiple times.

  Put it right: Assign the random value to a variable so the same value can be checked in multiple conditions.

• Confusing the index of an array with the random value itself.

  Put it right: Generate a random integer within the bounds of the array's indices (e.g. 0 to length-1) and use it to access the element.

Exam tips

• When writing pseudocode or code, clearly state the bounds for your random number generator, ensuring they match the scenario (e.g. 1 to 6 for a die).
• When a question asks for a simulation, state the range, how the value is stored in a variable, and how it is used in the program's logic.
• Check which language or pseudocode the question uses: OCR Reference Language `random(1, 6)` and Python `random.randint(1, 6)` both give a die roll but use different syntax.