randomChoice<T> function

T randomChoice<T>(Iterable<T> options, [ Iterable<double> weights = const [] ])

Selects a random element from options with optional weights.

If weights is provided, it will be used to select an option according to a weighted distribution. If it is not provided, an option will be selected according to a uniform distribution.

Throws an ArgumentError if options is empty. Throws an ArgumentError if weights is non-empty and does not match the length of options.

randomChoice(['hi', 'hello'], [0.7, 0.3]) // 70% chance of 'hi', 30% chance of 'hello'
randomChoice(['hi', 'hey']) // 'hi' and 'hey' are equally likely. Equivalent to randomChoice(['hi', 'hey'], [1, 1])

Implementation

T randomChoice<T>(Iterable<T> options, [Iterable<double> weights = const []]) {
  if (options.isEmpty) {
    throw ArgumentError.value(
        options.toString(), 'options', 'must be non-empty');
  }
  if (weights.isNotEmpty && options.length != weights.length) {
    throw ArgumentError.value(weights.toString(), 'weights',
        'must be empty or match length of options');
  }

  if (weights.isEmpty) {
    return options.elementAt(Random().nextInt(options.length));
  }

  double sum = weights.reduce((val, curr) => val + curr);
  double randomWeight = Random().nextDouble() * sum;

  int i = 0;
  for (int l = options.length; i < l; i++) {
    randomWeight -= weights.elementAt(i);
    if (randomWeight <= 0) {
      break;
    }
  }

  return options.elementAt(i);
}