Bobby/stock_system
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: Handle NaN ATR values and adjust lookback period in signal check

Browse Source
This commit is contained in:
Bobby (aider) 2025-02-19 22:28:23 -08:00
parent 96b4d951b4
commit 0a7d1afcde
1 changed files with 13 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/screener/t_candlestick.py
View File

@ -44,11 +44,14 @@ def check_entry_signal(df: pd.DataFrame, selected_patterns: list = None) -> list
Returns: Returns:
list: List of tuples (signal, date, signal_data) for each signal found list: List of tuples (signal, date, signal_data) for each signal found
""" """
if len(df) < 5: # Need at least 5 bars for pattern recognition if len(df) < 14: # Need at least 14 bars for ATR calculation
return [] return []
signals = [] signals = []
# Calculate ATR first
atr = talib.ATR(df['high'].values, df['low'].values, df['close'].values, timeperiod=14)
# Use selected patterns or all patterns if none selected # Use selected patterns or all patterns if none selected
patterns_to_scan = {k: v for k, v in CANDLESTICK_PATTERNS.items() patterns_to_scan = {k: v for k, v in CANDLESTICK_PATTERNS.items()
if selected_patterns is None or k in selected_patterns} if selected_patterns is None or k in selected_patterns}
@ -64,7 +67,7 @@ def check_entry_signal(df: pd.DataFrame, selected_patterns: list = None) -> list
) )
# Look for signals across all candles # Look for signals across all candles
for i in range(len(df)): for i in range(14, len(df)): # Start after ATR lookback period
found_patterns = [] found_patterns = []
for pattern_name, pattern_values in pattern_signals.items(): for pattern_name, pattern_values in pattern_signals.items():
@ -76,11 +79,15 @@ def check_entry_signal(df: pd.DataFrame, selected_patterns: list = None) -> list
# Calculate basic target and stop levels using recent price action # Calculate basic target and stop levels using recent price action
current_price = df.iloc[i]['close'] current_price = df.iloc[i]['close']
recent_low = df.iloc[max(0, i-5):i+1]['low'].min() # Look back 5 bars for stop recent_low = df.iloc[max(0, i-5):i+1]['low'].min() # Look back 5 bars for stop
atr = talib.ATR(df['high'].values, df['low'].values, df['close'].values, timeperiod=14)
# Use 2x ATR for target and 1x ATR for stop # Default to percentage-based calculations
target_price = current_price + (2 * atr[i] if not pd.isna(atr[i]) else current_price * 0.02) atr_value = atr[i]
stop_loss = max(recent_low, current_price - (atr[i] if not pd.isna(atr[i]) else current_price * 0.01)) if pd.isna(atr_value):
target_price = current_price * 1.02 # 2% target
stop_loss = current_price * 0.99 # 1% stop
else:
target_price = current_price + (2 * atr_value)
stop_loss = max(recent_low, current_price - atr_value)
signal_data = { signal_data = {
'price': current_price, 'price': current_price,