TAlib.Indicators.Stochastic (TAlib v0.13.0)

Copy Markdown View Source

Stochastic oscillator Wikipedia.

  • %K = (Close - Lowest Low) / (Highest High - Lowest Low) * 100
  • %D = SMA(%K, d_period)

See TAlib for the input conventions: prices are always given in chronological order, oldest first, and series are returned aligned with the input. nil is returned when there is not enough data or the window has no range (flat prices), since %K is then undefined.

Note that with a plain price list, close, high and low all come from the same series; for proper high/low support an OHLC input shape is needed.

Summary

Functions

Warm up streaming fast-%K state from history.

Advance streaming fast-%K state by one price, in O(1). A flat window (no range) yields nil, like the batch series.

The slow stochastic (TA-Lib STOCH): the slowed %K line is the SMA of the fast %K over slowing bars, and %D is the SMA of the slowed %K.

Warm up streaming slow-stochastic state (TA-Lib STOCH) from history: fast %K over period, slowed by an SMA of slowing, %D an SMA of d_period over the slowed line.

Advance streaming slow-stochastic state by one price, in O(1): {state, %{k: , d: }}.

%D: the d_period-period simple moving average of %K.

%K = (Current Close - Lowest Low) / (Highest High - Lowest Low) * 100

Calculate the full %K series.

Functions

k_init(prices, period \\ 14)

@spec k_init([number()], pos_integer()) ::
  {TAlib.Indicators.Stochastic.KState.t(), [float() | nil]}

Warm up streaming fast-%K state from history.

Returns {state, values} with values the full %K series of history (identical to stochastic_k_list/2), and state ready to continue it one price at a time in O(1) via k_next/2. History needs at least period prices.

Examples

iex> {state, values} = TAlib.Indicators.Stochastic.k_init([1, 6, 2, 5, 3], 3)
iex> values
[nil, nil, 20.0, 75.0, 33.33333333333333]
iex> {_state, 100.0} = TAlib.Indicators.Stochastic.k_next(state, 7)

k_next(state, price)

@spec k_next(TAlib.Indicators.Stochastic.KState.t(), number()) ::
  {TAlib.Indicators.Stochastic.KState.t(), float() | nil}

Advance streaming fast-%K state by one price, in O(1). A flat window (no range) yields nil, like the batch series.

iex> {state, _} = TAlib.Indicators.Stochastic.k_init([1, 6, 2, 5, 3], 3)
iex> {_state, value} = TAlib.Indicators.Stochastic.k_next(state, 7)
iex> value
100.0

slow(prices, period \\ 14, slowing \\ 3, d_period \\ 3)

@spec slow([number()], pos_integer(), pos_integer(), pos_integer()) :: %{
  k: float() | nil,
  d: float() | nil
}

The slow stochastic (TA-Lib STOCH): the slowed %K line is the SMA of the fast %K over slowing bars, and %D is the SMA of the slowed %K.

Parameters

  • prices: list of prices, oldest first.
  • period: %K lookback window, a positive integer (default 14).
  • slowing: SMA period applied to fast %K (default 3).
  • d_period: SMA period applied to the slowed %K (default 3).

Examples

iex> prices = [1, 6, 2, 5, 3, 7, 4]
iex> %{k: k, d: d} = TAlib.Indicators.Stochastic.slow(prices, 3, 2, 2)
iex> {Float.round(k, 4), Float.round(d, 4)}
{62.5, 64.5833}

slow_init(prices, period \\ 14, slowing \\ 3, d_period \\ 3)

@spec slow_init([number()], pos_integer(), pos_integer(), pos_integer()) ::
  {TAlib.Indicators.Stochastic.SlowState.t(),
   %{k: [float() | nil], d: [float() | nil]}}

Warm up streaming slow-stochastic state (TA-Lib STOCH) from history: fast %K over period, slowed by an SMA of slowing, %D an SMA of d_period over the slowed line.

Returns {state, values} with values the two aligned series (%{k: , d: }), and state ready to continue them one price at a time in O(1) via slow_next/2. History needs at least period + slowing + d_period - 2 prices.

Examples

iex> {state, values} = TAlib.Indicators.Stochastic.slow_init([1, 6, 2, 5, 3, 7, 4], 3, 2, 2)
iex> {List.last(values.k), List.last(values.d)}
{62.49999999999999, 64.58333333333333}
iex> {_state, %{k: k, d: d}} = TAlib.Indicators.Stochastic.slow_next(state, 5)
iex> {Float.round(k, 4), Float.round(d, 4)}
{29.1667, 45.8333}

slow_next(state, price)

@spec slow_next(TAlib.Indicators.Stochastic.SlowState.t(), number()) ::
  {TAlib.Indicators.Stochastic.SlowState.t(),
   %{k: float() | nil, d: float() | nil}}

Advance streaming slow-stochastic state by one price, in O(1): {state, %{k: , d: }}.

iex> {state, _} = TAlib.Indicators.Stochastic.slow_init([1, 6, 2, 5, 3, 7, 4], 3, 2, 2)
iex> {_state, %{k: k, d: d}} = TAlib.Indicators.Stochastic.slow_next(state, 5)
iex> {Float.round(k, 4), Float.round(d, 4)}
{29.1667, 45.8333}

stochastic_d(prices, period \\ 14, d_period \\ 3)

@spec stochastic_d([number()], pos_integer(), pos_integer()) :: float() | nil

%D: the d_period-period simple moving average of %K.

Parameters

  • prices: list of prices, oldest first.
  • period: %K lookback window, a positive integer (default 14).
  • d_period: SMA period applied to %K, a positive integer (default 3).

Examples

iex> TAlib.Indicators.Stochastic.stochastic_d([1, 6, 2, 5, 3, 7, 4], 5, 3)
60.0

stochastic_k(prices, period \\ 14)

@spec stochastic_k([number()], pos_integer()) :: float() | nil

%K = (Current Close - Lowest Low) / (Highest High - Lowest Low) * 100

Parameters

  • prices: list of prices, oldest first.
  • period: lookback window, a positive integer (default 14).

Examples

iex> TAlib.Indicators.Stochastic.stochastic_k([1, 2, 5, 1, 4], 5)
75.0

iex> TAlib.Indicators.Stochastic.stochastic_k([5, 5, 5], 3)
nil

stochastic_k_list(prices, period \\ 14)

@spec stochastic_k_list([number()], pos_integer()) :: [float() | nil]

Calculate the full %K series.

Returns a list aligned with prices: the first period - 1 entries are nil, and every later entry is %K of the window ending at that point. Windows with no range (flat prices) produce nil, since %K is then undefined.

Parameters

  • prices: list of prices, oldest first.
  • period: lookback window, a positive integer (default 14).

Examples

iex> TAlib.Indicators.Stochastic.stochastic_k_list([1, 6, 2, 5, 3], 3)
[nil, nil, 20.0, 75.0, 33.33333333333333]