Developing Indicators & Strategy Logic¶

Guide to testing indicator expressions and iterating on strategy logic.

Indicator Playground¶

Test indicator calculations without running a full backtest.

Quick Indicator Test¶

tradai indicator test "ta.RSI(df['close'], 14)"

Output:

Testing indicator on BTC/USDT:USDT
Period: 2024-11-22 to 2024-12-22

┌──────────────────────────────────────────────────┐
│                    RSI Stats                      │
├────────────┬─────────────────────────────────────┤
│ Count      │ 720                                 │
│ Min        │ 23.41                               │
│ Max        │ 78.23                               │
│ Mean       │ 52.18                               │
│ Current    │ 45.67                               │
└────────────┴─────────────────────────────────────┘

Last 50 values: ▃▄▅▆▇▇▆▅▄▃▂▃▄▅▆▇▆▅▄▃▂▁▂▃▄▅▆▅▄▃▂▃▄▅▆▇▆▅▄▃▂▃▄▅▆▇▆▅

Test with Different Symbols/Timeframes¶

# Different symbol
tradai indicator test "ta.RSI(df['close'], 14)" --symbol ETH/USDT:USDT

# Different timeframe
tradai indicator test "ta.MACD(df['close'])" --timeframe 4h

# Custom period
tradai indicator test "ta.EMA(df['close'], 20)" --period 7d

Test Complex Expressions¶

# MACD histogram
tradai indicator test "ta.MACD(df['close'])[2]"

# Bollinger Bands width
tradai indicator test "(ta.BBANDS(df['close'])[0] - ta.BBANDS(df['close'])[2]) / ta.BBANDS(df['close'])[1]"

# Custom calculations
tradai indicator test "df['close'].pct_change().rolling(20).std()"

Use Different Data Sources¶

# From Binance (default)
tradai indicator test "ta.RSI(df['close'])" --source binance

# From ArcticDB (if data synced)
tradai indicator test "ta.RSI(df['close'])" --source arctic

# From local file
tradai indicator test "ta.RSI(df['close'])" --source file --file ./data.csv

Available Indicators¶

View all available TA-Lib indicators:

tradai indicator list

Output shows categorized indicators:

Trend Indicators¶

Indicator	Expression	Description
SMA	`ta.SMA(df['close'], 20)`	Simple Moving Average
EMA	`ta.EMA(df['close'], 20)`	Exponential Moving Average
MACD	`ta.MACD(df['close'])`	MACD (macd, signal, hist)
ADX	`ta.ADX(df['high'], df['low'], df['close'])`	Average Directional Index

Momentum Indicators¶

Indicator	Expression	Description
RSI	`ta.RSI(df['close'], 14)`	Relative Strength Index
STOCH	`ta.STOCH(df['high'], df['low'], df['close'])`	Stochastic (slowk, slowd)
CCI	`ta.CCI(df['high'], df['low'], df['close'])`	Commodity Channel Index
MOM	`ta.MOM(df['close'], 10)`	Momentum

Volatility Indicators¶

Indicator	Expression	Description
BBANDS	`ta.BBANDS(df['close'])`	Bollinger Bands (upper, mid, lower)
ATR	`ta.ATR(df['high'], df['low'], df['close'])`	Average True Range
NATR	`ta.NATR(df['high'], df['low'], df['close'])`	Normalized ATR

Volume Indicators¶

Indicator	Expression	Description
OBV	`ta.OBV(df['close'], df['volume'])`	On Balance Volume
MFI	`ta.MFI(df['high'], df['low'], df['close'], df['volume'])`	Money Flow Index

Interactive REPL¶

For exploratory analysis, use the interactive REPL:

tradai indicator repl

This opens an IPython session with pre-loaded data:

# Data already loaded as 'df'
>>> df.tail()
                          date      open      high       low     close    volume
2024-12-22 09:00:00  97500.12  98200.45  97400.00  97850.23  1234567.8

# TA-Lib available as 'ta'
>>> rsi = ta.RSI(df['close'], 14)
>>> rsi.tail()

# Pandas/NumPy available
>>> df['returns'] = df['close'].pct_change()
>>> df['volatility'] = df['returns'].rolling(20).std()

Syncing Market Data¶

Before testing indicators, ensure you have data synced:

# Check data availability
tradai data check-freshness BTC/USDT:USDT

# Sync data if needed
tradai data sync BTC/USDT:USDT \
  --start 2024-01-01 \
  --end 2024-12-22 \
  --timeframe 1h

Strategy Linting¶

Check your strategy for common issues:

tradai strategy lint MyStrategy

Detected Issues¶

Look-Ahead Bias

strategies/my_strategy/strategy.py:45: error: Potential look-ahead bias
  df.shift(-1) accesses future data

Missing Warmup Period

strategies/my_strategy/strategy.py:1: warning: Missing startup_candle_count
  Strategy uses indicators but doesn't set startup_candle_count

Invalid Timeframe/Category Combination

warning: Scalping strategies typically use short timeframes (1m, 5m, 15m)
  Current: 4h

Development Workflow¶

1. Write Indicator Code¶

def populate_indicators(self, dataframe, metadata):
    dataframe["rsi"] = ta.RSI(dataframe["close"], timeperiod=14)
    return dataframe

2. Test Indicator Quickly¶

tradai indicator test "ta.RSI(df['close'], 14)" --symbol BTC/USDT:USDT

3. Lint Strategy¶

tradai strategy lint MyStrategy

4. Run Unit Tests¶

just test-package strategies/my_strategy

5. Quick Backtest¶

tradai backtest quick MyStrategy --period 7d

Common Patterns¶

RSI Oversold/Overbought¶

# Entry: RSI < 30 (oversold)
dataframe.loc[dataframe["rsi"] < 30, "enter_long"] = 1

# Exit: RSI > 70 (overbought)
dataframe.loc[dataframe["rsi"] > 70, "exit_long"] = 1

Moving Average Crossover¶

# Entry: Fast EMA crosses above slow EMA
dataframe.loc[
    (dataframe["ema_fast"] > dataframe["ema_slow"]) &
    (dataframe["ema_fast"].shift(1) <= dataframe["ema_slow"].shift(1)),
    "enter_long"
] = 1

Bollinger Band Squeeze¶

# Entry: Price below lower band
lower_band = ta.BBANDS(dataframe["close"])[2]
dataframe.loc[dataframe["close"] < lower_band, "enter_long"] = 1

Quick Reference¶

Command	Description
`tradai indicator test EXPR`	Test indicator expression
`tradai indicator list`	List all indicators
`tradai indicator repl`	Interactive REPL
`tradai strategy lint NAME`	Check for issues
`tradai data check-freshness SYMBOL`	Check data availability
`tradai data sync SYMBOL`	Sync market data