bitmex_and_binance.py

# IMPORTS
import pandas as pd
import math
import os.path
import time
from bitmex import bitmex
from binance.client import Client
from datetime import timedelta, datetime
from dateutil import parser
from tqdm import tqdm_notebook #(Optional, used for progress-bars)

### API
bitmex_api_key = '[REDACTED]'    #Enter your own API-key here
bitmex_api_secret = '[REDACTED]' #Enter your own API-secret here
binance_api_key = '[REDACTED]'    #Enter your own API-key here
binance_api_secret = '[REDACTED]' #Enter your own API-secret here

### CONSTANTS
binsizes = {"1m": 1, "5m": 5, "1h": 60, "1d": 1440}
batch_size = 750
bitmex_client = bitmex(test=False, api_key=bitmex_api_key, api_secret=bitmex_api_secret)
binance_client = Client(api_key=binance_api_key, api_secret=binance_api_secret)


### FUNCTIONS
def minutes_of_new_data(symbol, kline_size, data, source):
    if len(data) > 0:  old = parser.parse(data["timestamp"].iloc[-1])
    elif source == "binance": old = datetime.strptime('1 Jan 2017', '%d %b %Y')
    elif source == "bitmex": old = bitmex_client.Trade.Trade_getBucketed(symbol=symbol, binSize=kline_size, count=1, reverse=False).result()[0][0]['timestamp']
    if source == "binance": new = pd.to_datetime(binance_client.get_klines(symbol=symbol, interval=kline_size)[-1][0], unit='ms')
    if source == "bitmex": new = bitmex_client.Trade.Trade_getBucketed(symbol=symbol, binSize=kline_size, count=1, reverse=True).result()[0][0]['timestamp']
    return old, new

def get_all_binance(symbol, kline_size, save = False):
    filename = '%s-%s-data.csv' % (symbol, kline_size)
    if os.path.isfile(filename): data_df = pd.read_csv(filename)
    else: data_df = pd.DataFrame()
    oldest_point, newest_point = minutes_of_new_data(symbol, kline_size, data_df, source = "binance")
    delta_min = (newest_point - oldest_point).total_seconds()/60
    available_data = math.ceil(delta_min/binsizes[kline_size])
    if oldest_point == datetime.strptime('1 Jan 2017', '%d %b %Y'): print('Downloading all available %s data for %s. Be patient..!' % (kline_size, symbol))
    else: print('Downloading %d minutes of new data available for %s, i.e. %d instances of %s data.' % (delta_min, symbol, available_data, kline_size))
    klines = binance_client.get_historical_klines(symbol, kline_size, oldest_point.strftime("%d %b %Y %H:%M:%S"), newest_point.strftime("%d %b %Y %H:%M:%S"))
    data = pd.DataFrame(klines, columns = ['timestamp', 'open', 'high', 'low', 'close', 'volume', 'close_time', 'quote_av', 'trades', 'tb_base_av', 'tb_quote_av', 'ignore' ])
    data['timestamp'] = pd.to_datetime(data['timestamp'], unit='ms')
    if len(data_df) > 0:
        temp_df = pd.DataFrame(data)
        data_df = data_df.append(temp_df)
    else: data_df = data
    data_df.set_index('timestamp', inplace=True)
    if save: data_df.to_csv(filename)
    print('All caught up..!')
    return data_df

def get_all_bitmex(symbol, kline_size, save = False):
    filename = '%s-%s-data.csv' % (symbol, kline_size)
    if os.path.isfile(filename): data_df = pd.read_csv(filename)
    else: data_df = pd.DataFrame()
    oldest_point, newest_point = minutes_of_new_data(symbol, kline_size, data_df, source = "bitmex")
    delta_min = (newest_point - oldest_point).total_seconds()/60
    available_data = math.ceil(delta_min/binsizes[kline_size])
    rounds = math.ceil(available_data / batch_size)
    if rounds > 0:
        print('Downloading %d minutes of new data available for %s, i.e. %d instances of %s data in %d rounds.' % (delta_min, symbol, available_data, kline_size, rounds))
        for round_num in tqdm_notebook(range(rounds)):
            time.sleep(1)
            new_time = (oldest_point + timedelta(minutes = round_num * batch_size * binsizes[kline_size]))
            data = bitmex_client.Trade.Trade_getBucketed(symbol=symbol, binSize=kline_size, count=batch_size, startTime = new_time).result()[0]
            temp_df = pd.DataFrame(data)
            data_df = data_df.append(temp_df)
    data_df.set_index('timestamp', inplace=True)
    if save and rounds > 0: data_df.to_csv(filename)
    print('All caught up..!')
    return data_df

Getting this annoying message

C:\Users\Paul\anaconda3\lib\site-packages\dateparser\date_parser.py:35: PytzUsageWarning: The localize method is no longer necessary, as this time zone supports the fold attribute (PEP 495). For more details on migrating to a PEP 495-compliant implementation, see https://pytz-deprecation-shim.readthedocs.io/en/latest/migration.html
date_obj = stz.localize(date_obj)

Thanks Bro, its amazing