@spec fibonacci_sequence(pos_integer(), pos_integer()) :: [pos_integer()]

Generate Fibonacci sequence within a range.

Useful for parameter values that benefit from Fibonacci spacing.

Examples

iex> fibonacci_sequence(1, 100)
[1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89]

@spec linspace(number(), number(), pos_integer()) :: [float()]

Generate linearly spaced values similar to numpy.linspace.

Examples

iex> linspace(0, 10, 5)
[0.0, 2.5, 5.0, 7.5, 10.0]

@spec logarithmic_range(pos_integer(), pos_integer(), pos_integer()) :: [
  pos_integer()
]

Generate logarithmic range.

Creates exponentially spaced values, useful for parameters like periods or thresholds that work better on a log scale.

Examples

iex> logarithmic_range(1, 64, 2)
[1, 2, 4, 8, 16, 32]

@spec parameter_grid(%{required(atom()) => Range.t() | [any()] | any()}) ::
  {:ok, [%{required(atom()) => any()}]} | {:error, term()}

Generate all combinations of parameters from a parameter map.

Takes a map where keys are parameter names and values are either:

• Ranges (e.g., 1..10)
• Lists of specific values (e.g., [5, 10, 15, 20])
• Single values (treated as a list with one element)

Examples

iex> param_map = %{fast_period: 5..7, slow_period: [20, 25]}
iex> {:ok, combinations} = parameter_grid(param_map)
iex> length(combinations)
6
iex> Enum.all?(combinations, &Map.has_key?(&1, :fast_period))
true

@spec random_search(%{required(atom()) => Range.t() | [any()]}, pos_integer()) ::
  {:ok, [%{required(atom()) => any()}]} | {:error, term()}

Generate random parameter combinations.

Useful for large parameter spaces where exhaustive search is impractical.

Examples

iex> param_map = %{fast_period: 5..50, slow_period: 20..100}
iex> {:ok, samples} = random_search(param_map, 10)
iex> length(samples)
10

@spec range(number(), number(), number()) :: [number()]

Generate a range similar to numpy.arange.

Examples

iex> range(1, 5)
[1, 2, 3, 4]

iex> range(0, 10, 2)
[0, 2, 4, 6, 8]