gpt-4-0314· OpenAI Arena Human Preference 55K
write hummingbot script, buy and sell following fast and slow MA cross
- User
write hummingbot script, buy and sell following fast and slow MA cross
- Assistant
Below is a basic Hummingbot script that buys when a fast Moving Average (MA) crosses above the slow Moving Average and sells when it crosses below.
#################################### ## My Hummingbot MA Strategy ## #################################### # Load necessary libraries import numpy as np # Configure strategy parameters fast_ma_len = 5 slow_ma_len = 15 min_cross_threshold = 0.01 # Calculate moving averages def calculate_moving_average(data, _len): return data[-_len:].mean() # Define cross detect function def detect_cross(fast_ma, slow_ma, cross_threshold): if fast_ma <= slow_ma * (1 + cross_threshold): return 'buy' elif fast_ma >= slow_ma * (1 - cross_threshold): return 'sell' else: return 'hold' # Get exchange and trading pair exchange = get_active_exchange() trading_pair = exchange.trading_pair # Get trading pair's price history price_history = get_mid_price_history(exchange=exchange, trading_pair=trading_pair) # Create a numpy array with price history price_history_np = np.array(price_history) # Calculate moving averages fast_ma = calculate_moving_average(price_history_np, fast_ma_len) slow_ma = calculate_moving_average(price_history_np, slow_ma_len) # Detect cross signal = detect_cross(fast_ma, slow_ma, min_cross_threshold) # Execute strategy if signal == 'buy': execute_buy(exchange=exchange, trading_pair=trading_pair) elif signal == 'sell': execute_sell(exchange=exchange, trading_pair=trading_pair) else: passThis script calculates the fast and slow MAs from the provided price history for the specified lengths and checks for the MA cross. If a valid cross exists, the script will execute a buy or sell order accordingly. Please make sure to replace the placeholder functions like
execute_buy,execute_sell,get_active_exchange, andget_mid_price_historywith actual Hummingbot functionalities.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.