Hi @jeff.kenyon

If you want to consume tick data after the realtime event then it is classed as historical data and will, therefore, require the use of the historical data source and related functions. The lags will be dependent on the underlying data source. My colleague has reached out to the API Product owner to ask him to comment on the lags etc.

As the Refintiv Data Platform library is in Beta, full documentation is not yet available - but will be added in due course. The Product owner should again be able to advise on this.

If you want the data in the most timely manner then you can use the Streaming Price interface in event-driven mode - and thereby avoid having to make constant snapshot request. i.e. open the request once, handle all the responses for the require duration and then close the stream once done.

So, you could Open the StreamingPrice instance at say 08:00 and close it at 8:20

Two examples you can refer to:

Example 2.3.1 Streaming Price with all Events

Example 2.3.0 - Update Dataframe

Note that when you connect to ERT in Cloud i.e. Platform session, the feed is Trade Safe conflated - i.e. up to 3 Quotes events a second + all Trade events.

If however, you are using a Deployed session - then you will need to check with your MarketData team that you are using a full tick Service - if that is what you are after.

You can optionally specify the service name (for a deployed session) as follows:

prices = rdp.StreamingPrices(session = mySession, 
    service = 'ELEKTRON_EDGE',
    universe = ['EUR=','CHF=', 'AED='],
    fields = ['BID', 'ASK','DSPLY_NAME', 'QUOTIM'])

If you don't specify a service name, then it will use whichever service your MarketData team have specified as the default service for Websocket type connections.

Hi @Umer.Nalla,

Thanks very much for the detailed response. After getting permissioned for the data I need, I used the example you cited, and the ticks I need are now showing up in my Jupyter notebook. Final question: what I need to do is collect the ticks to create a CSV file for the time window of interest? Options seem to be to collect everything in a data frame, then write out the data frame after I close the streaming prices, or write each tick to a file as I go. Unfortunately, I can't seem to do either at the moment. Are there any examples of writing the tick data to a file?

Note that I can't just use an ending summary...I need all the individual ticks, which I'll then be doing some analysis on in another program.

Thanks!

An addedum to the above...I've defined my handlers for the refresh, update, and status events (below), and it runs great in a Jupyter notebook: I have a file handle f that is opened before the RDP stream, and flushed and closed after. I get my desired CSV file of events. When I use the Jupyter notebook code in a standard module, and run it from the Windows command line...nothing happens. The StreamState is opened (I checked), I wait for n seconds, then close the stream and the session, but my destination file is empty. It seems as if no events are coming through on the open stream.

def display_refreshed_fields(streaming_price, instrument_name, fields):
    current_time = datetime.datetime.now()
    timestampStr = current_time.strftime("%Y-%m-%d %H:%M:%S.%f")
    output_string = timestampStr + ",REFRESH," + instrument_name + ",\"" + str(fields) + "\"\n"
    f.write(output_string)
    print(current_time, "- Refresh received for", instrument_name, ":", fields)
    
def display_updated_fields(streaming_price, instrument_name, fields):
    current_time = datetime.datetime.now()
    timestampStr = current_time.strftime("%Y-%m-%d %H:%M:%S.%f")
    output_string = timestampStr + ",UPDATE," + instrument_name + ",\"" + str(fields) + "\"\n"
    f.write(output_string)
    print(current_time, "- Update received for", instrument_name, ":", fields)
    
def display_status(streaming_price, instrument_name, status):
    current_time = datetime.datetime.now()
    timestampStr = current_time.strftime("%Y-%m-%d %H:%M:%S.%f")
    output_string = timestampStr + ",STATUS," + instrument_name + ",\"" + str(status) + "\"\n"
    f.write(output_string)
    print(current_time, "- Status received for", instrument_name, ":", status)

Anyway...very puzzled as to why this works perfectly in Jupyter Notebook, and not at all from console. Suggestions welcome.

Hi @jeff.kenyon

Apologies that I did not get back to you first follow up question of the Apr 15th - seems there was a problem with (my?) email notifications and I was not aware of the follow-up.

So, the current situation is that you managed to answer that question for yourself but are now stuck when you run the code as a script rather than a notebook.

If you can share the full script -minus any username etc then I would be happy to try it out here and see if I can recreate your issue.

TestStreamingScript.txt

About to open!
Opened! 
StreamState.Open

Hi @jeff.kenyon,

One thing I noticed when using the streaming interfaces within a notebook vs running from a command line is that within a Jupyter Notebook, an event loop is automatically created for you. Try the following as the last line within your main when running from a command line:

asyncio.get_event_loop().run_forever() 

Hi @jeff.kenyon

Ok - this had me scratching my head for several minutes - my near-identical script worked fine - but the one you supplied did not.

Line by line comparison showed it turned out to be a typo ...

universe = ['EUR=', 'GBP=', 'JPY=', 'CAD='] 
fields   = ['BID', 'ASK', 'OPEN_PRC', 'HST_CLOSE']

If you remove the "" from the universe and fields variable definitions so that you are defining a list rather than a string it should work.

Oh, and you will need to change the sleep to something like the following:

end_time = time.time() + WINDOW_LENGTH 
while (time.time() < end_time):      
    asyncio.get_event_loop().run_until_complete(asyncio.sleep(1)) 
streaming_prices.close() 

Otherwise, the Update handlers will never trigger.

A modified version of your script: teststreamingscriptx.txt