Advanced PyPScript Techniques

Advanced PyPScript is not about making expressions longer.

It is about making strategies more intentional:

• cleaner regime filters
• better long and short symmetry, or intentional asymmetry
• more disciplined confidence ladders
• better reuse of conditions

1. Layered Confirmation

A strong pattern usually has layers:

• regime
• direction
• trigger

Example:

trend_ma = sma.evaluate(close, 50)
entry_ma = sma.evaluate(close, 20)
rsi_value = rsi.evaluate(close, 14)

trend_up = above(close, trend_ma)
momentum_up = above(rsi_value, 55)
entry_reclaim = cross(close, entry_ma)

patterns {
    trend_up and momentum_up and entry_reclaim -> UP(confidence: 0.86)
    default -> HOLD(confidence: 0.95)
}

Each line has a job:

• trend_up says the market backdrop is acceptable
• momentum_up says the move has internal support
• entry_reclaim says the timing is here now

2. Regime Filters

Many bad strategies fail because they trade every environment.

Use regime filters to say when a setup is even allowed.

atr_value = atr.evaluate(high, low, close, 14)
volatility_ok = above(atr_value, 0.0)

patterns {
    trend_up and momentum_up and entry_reclaim and volatility_ok -> UP(confidence: 0.85)
    default -> HOLD(confidence: 0.96)
}

A regime filter should do one thing clearly:

• trend only
• volatility only
• momentum only

Do not overload one variable with three meanings.

3. Asymmetric Long And Short Logic

Markets often do not behave symmetrically.

If long and short quality differ, reflect that directly in the script.

patterns {
    trend_up and long_momentum and long_trigger -> UP(confidence: 0.80)
    trend_down and short_momentum and short_trigger -> DOWN(confidence: 0.90)
    default -> HOLD(confidence: 0.96)
}

That is not inconsistency.

That is honesty.

4. Confidence Ladders

Instead of one giant “all-or-nothing” rule, use a confidence ladder.

patterns {
    trend_up and momentum_up and breakout and volatility_ok -> UP(confidence: 0.90)
    trend_up and momentum_up and breakout -> UP(confidence: 0.82)
    trend_up and breakout -> UP(confidence: 0.72)
    default -> HOLD(confidence: 0.96)
}

This works well when:

• you want the same directional idea
• but with different levels of confirmation

5. Explicit Weak-Zone Holds

Advanced scripts are often better because they are better at refusing bad states.

momentum_mid = between(rsi_value, 45, 55)

patterns {
    momentum_mid -> HOLD(confidence: 0.97)
    trend_up and breakout -> UP(confidence: 0.82)
    trend_down and breakdown -> DOWN(confidence: 0.82)
    default -> HOLD(confidence: 0.95)
}

That can be more powerful than adding another aggressive entry rule.

6. Compose Conditions, Do Not Repeat Them

Advanced scripts stay readable by building reusable components.

trend_up = above(fast_ma, slow_ma)
trend_down = below(fast_ma, slow_ma)
volatility_ok = above(atr_value, 0.0)
long_context = trend_up and volatility_ok
short_context = trend_down and volatility_ok

patterns {
    long_context and long_trigger -> UP(confidence: 0.84)
    short_context and short_trigger -> DOWN(confidence: 0.84)
    default -> HOLD(confidence: 0.96)
}

7. Use Thresholds Intentionally

Thresholds should mean something.

Weak:

above(rsi_value, 51)

unless you know exactly why 51 matters.

Better:

above(rsi_value, 55)
below(rsi_value, 45)
between(rsi_value, 45, 55)

Those values clearly define bullish, bearish, and neutral zones.

8. Advanced Example

#!pyp/1.0
engine: pyp-1.0

import indicators.sma
import indicators.rsi
import indicators.atr
import functions.above
import functions.below
import functions.between
import functions.cross

strategy "advanced_regime_breakout" {
    fast_ma = sma.evaluate(close, 20)
    slow_ma = sma.evaluate(close, 50)
    rsi_value = rsi.evaluate(close, 14)
    atr_value = atr.evaluate(high, low, close, 14)

    trend_up = above(fast_ma, slow_ma)
    trend_down = below(fast_ma, slow_ma)
    long_momentum = above(rsi_value, 58)
    short_momentum = below(rsi_value, 42)
    neutral_momentum = between(rsi_value, 45, 55)
    volatility_ok = above(atr_value, 0.0)
    breakout_long = cross(close, fast_ma)
    breakout_short = cross(fast_ma, close)

    long_context = trend_up and long_momentum and volatility_ok
    short_context = trend_down and short_momentum and volatility_ok

    patterns {
        neutral_momentum -> HOLD(confidence: 0.98)
        long_context and breakout_long -> UP(confidence: 0.88)
        short_context and breakout_short -> DOWN(confidence: 0.90)
        trend_up and breakout_long -> UP(confidence: 0.74)
        trend_down and breakout_short -> DOWN(confidence: 0.78)
        default -> HOLD(confidence: 0.96)
    }
}

What “Advanced” Should Mean

Advanced should mean:

• clearer logic under more complex market conditions
• better selectivity
• more truthful confidence

It should not mean:

• unreadable one-liners
• five indicators repeated everywhere
• patterns nobody can explain

Related Pages

• Pattern Matching
• Best Practices
• Troubleshooting