create basket algo

Author: GitHub-Valie

algorithms/live algos/basket_algo.py

@@ -0,0 +1,116 @@
+# Based on basic_live_algo.py, this algorithm's trading strategy 
+# is different in that it will take LONG / SHORT on a basket of 
+# technical indicators instead of a "simple" SMA strategy
+
+# Imports
+import json, requests, websocket
+from pprint import pprint
+import numpy as np
+
+
+# Define variables
+SYMBOL = "BTCUSDT"
+INTERVAL = "1m"
+WSS_BASE_URL = "wss://fstream.binance.com"
+WSS_ENDPOINT = "/ws/{}@kline_{}".format(
+    SYMBOL.lower(),
+    INTERVAL
+)
+data = []
+closes = []
+position = 0
+
+
+# Function to get historical data
+def futures_get_hist(symbol, interval):
+    '''
+    Get historical klines for a futures pair    
+    :symbol: str, i.e "BTCUSDT"     
+    :interval: str, i.e "1m"    
+    '''
+    r = requests.get(
+        "https://fapi.binance.com" + "/fapi/v1/klines",
+        params = {
+            'symbol': symbol,
+            'interval': interval
+        }
+    )
+
+    req = r.json()
+    return req
+
+
+# Technical indicators
+# Trend
+def sma(arr, window):
+    '''
+    Standard Moving Average     
+    :arr: array     
+    :window: window on which SMA will be calculated     
+    '''
+    return np.convolve(arr, np.ones(window), 'valid') / window
+
+
+def ewma(arr, alpha, window):
+    '''
+    Exponential Weighted Moving Average     
+    :arr: array     
+    :param alpha: specify decay [0, 1]      
+    :window: length of ewma 
+    '''
+    arr = arr[-(len(arr) - window + 1):] # Re-adjusting length of arr so it has the same len as sma
+    ewma_arr = np.zeros_like(arr) # returns an array of zeros the same length as arr
+    ewma_arr[0] = arr[0] # first value in list ewma_arr is equal to first value in list arr
+    for t in range(1, arr.shape[0]):
+        ewma_arr[t] = alpha * arr[t] + (1 - alpha) * ewma_arr[t - 1]
+    
+    return ewma_arr
+
+
+# Momentum
+def rsi():
+    return rsi
+
+def macd():
+    return macd
+
+# Volatility
+def bbands(arr, window):
+    '''
+    Bollinger Bands. Given an array and a certain timeframe (window), returns:      
+    `bb_mid` : simple moving average    
+    `bb_top` : bb_mid + 2 standard devs     
+    `bb_bot` : bb_mid - 2 standard devs     
+        
+    params:     
+    :arr: array, list object    
+    :window: int    
+    '''
+    bb_mid = sma(
+        arr,
+        window=window
+    )[-1]
+
+    bb_top = bb_mid + 2 * np.std(
+        arr,
+        ddof=1
+    )
+
+    bb_bot = bb_mid - 2 * np.std(
+        arr,
+        ddof=1
+    )
+
+    bollingerbands = [
+        bb_bot,
+        bb_mid,
+        bb_top
+    ]
+
+    return bollingerbands
+
+
+# Websocket functions
+
+
+# WebSocketApp