Question

我在使用python jupyter的websockets运行异步时遇到一些麻烦的问题。我读了这个现有的主题：How do I run Python asyncio code in a Jupyter notebook?

但是，当我应用刚刚运行的修复程序时，仍然会出错：在Jupyter python中等待。

当我使用标准代码行时：

asyncio.get_event_loop().run_until_complete(call_api(json.dumps(msg)))

我的错误消息如下：

RuntimeError                              Traceback (most recent call last)
<ipython-input-40-23500a349044> in <module>
     65 asyncio.run_coroutine_threadsafe(call_api(msg), loop)
     66 
---> 67 asyncio.get_event_loop().run_until_complete(call_api(json.dumps(msg)))
     68 

C:\ProgramData\Anaconda3\lib\asyncio\base_events.py in run_until_complete(self, future)
    569         future.add_done_callback(_run_until_complete_cb)
    570         try:
--> 571             self.run_forever()
    572         except:
    573             if new_task and future.done() and not future.cancelled():

C:\ProgramData\Anaconda3\lib\asyncio\base_events.py in run_forever(self)
    524         self._check_closed()
    525         if self.is_running():
--> 526             raise RuntimeError('This event loop is already running')
    527         if events._get_running_loop() is not None:
    528             raise RuntimeError(

RuntimeError: This event loop is already running

我将修复程序应用于参考线程，如下所示

# jupyter cannot use python main syntax and have to use loop form below.
loop = asyncio.get_event_loop()
loop.create_task(call_api(msg))
asyncio.run_coroutine_threadsafe(call_api(msg), loop)

但是，除了代码行如下：

<Future at 0xa02de10 state=pending>

完整代码如下。

# this code seems to work

import pandas as pd
import numpy as np
import scipy as sc
import datetime as dt
import time
import matplotlib.pyplot as plt

from datetime import date
import asyncio
import websockets
import json

from IPython.display import clear_output

# To subscribe to this channel:
msg = \
    {"jsonrpc": "2.0",
     "method": "public/subscribe",
     "id": 42,
     "params": {
        "channels": ["markprice.options.btc_usd"]}
    }


db = []
column = ['iv', 'mark_price', 'commod', 'expiry', 'strike', 'callput']

async def call_api(msg, t_delay =10, t_period =1):
    async with websockets.connect('wss://test.deribit.com/ws/api/v2') as websocket:
        await websocket.send(msg)

        c_time = 0.0 # initialize start time
        s_time = time.time() + t_delay # get end time time

        while (websocket.open)& (c_time <= s_time): # this is while webscoket is open, we use 'while websocket.open:'
            response = await websocket.recv()
            d = json.loads(response)
            if 'params' not in d:
                continue

            else:
                db = d['params']['data']

            # sets up the panda dataframe and the information transform.
            df = pd.DataFrame(db)
            df['splitinst'] = df['instrument_name'].str.split("-")
            df[['commod','expiry','strike','callput']] = pd.DataFrame(df.splitinst.values.tolist(), index=df.index)
            df['expiry'] = pd.to_datetime(df['expiry'])
            df = df.sort_values(by=['expiry','strike'])

            df.to_json('C:/settles.json', orient='records')
            df.to_csv('C:/settles.csv')


            await asyncio.sleep(t_period) # wait for this seconds before doing loo again
            c_time = time.time() # update the time

            clear_output(wait = True)
            print('current time is', c_time)

            print(df.iloc[0])

# jupyter cannot use python main syntax and have to use loop form below.
loop = asyncio.get_event_loop()
loop.create_task(call_api(msg))
asyncio.run_coroutine_threadsafe(call_api(msg), loop)

# asyncio.get_event_loop().run_until_complete(call_api(json.dumps(msg)))

在python Jupyter中调试asyncio websockets

0 个答案: