Indicators

The Indicators system in Buildalgos.com lets you define and cache technical indicators declaratively using applyIndicator(). You typically register them once in indicator_init() and then access the computed series via named attributes like self.<name>.

applyIndicator – Syntax & Parameters

def applyIndicator(
    self,
    indicatorName: str,
    *,
    name: str,
    **kwargs
) -> pandas.Series | pandas.DataFrame

Parameter	Required	Description
indicator Name	✔	Name of the indicator to compute.
name	✔	Attribute name to access later (self.<name>).
**kwargs	depends	Inputs specific to the indicator (many accept callables).
Return	depends	Returns value if called outside indicator_init(); otherwise stores to self.<name>.

Supported Indicators & Required Kwargs

Below is a complete list of supported indicators, their keyword arguments, and return types:

Indicator Name	Required Kwargs
sma / ema	close, length
rsi	close, length
macd	close, fast, slow, signal
vwap	df (OHLCV)
supertrend	high, low, close, length, multiplier
parabolicSAR	high, low, start, increment, maximum
cci	df, length, constant
bbands	close, length, stdDev, mode
stoch_oscillator	high, low, close, k, d, smooth_k
stoch_rsi	close, length, rsi_length, k, d
adx	high, low, close, length, length_signal
donchian_channels	high, low, length
williams_R	high, low, close, length
fisher_transform	high, low, length
hl2 / hlc3	df
prevDaysRange	df, currentDay, prevDays
pivotpoints	df, currentDay
orb	df, currentDay, length
heikenashi	df

Why Use lambda for Inputs?

Use callables (like lambda: self.dt_SPOT['close']) instead of static series so that:

• Data is evaluated on-demand, per-bar.
• You always get the up-to-date value from the right bar in live or backtest mode.
• It supports efficient caching and avoids unnecessary recomputation

Example:

def indicator_init(self):
    # Register 21-period EMA of close prices
    self.applyIndicator(
        indicatorName="ema",
        name="ema_21",
        close=lambda: self.dt_SPOT["close"],
        length=21
    )

    # Register MACD
    self.applyIndicator(
        indicatorName="macd",
        name="macd_main",
        close=lambda: self.dt_SPOT["close"],
        fast=12,
        slow=26,
        signal=9
    )

Now use:

if self.ema_21.iloc[-1] > self.dt_SPOT["close"].iloc[-1]:
    # your logic here

Custom Indicators

You can create custom indicators using only Pandas and NumPy to ensure maximum compatibility and performance with the framework.