In Python package "refinitiv-dataplatform", what is the the limit of allowed requests for "rdp.conve

I see it works for 600 but not for 10.000. What is the limit?

Best Answer

Answers

  • Hello @nicola.fiore

    I want to add more information about the Symbology functions document besides my colleague's answer.
    You can log in to the Refinitiv RDP APIs Document page with your RDP credential and get more detail about the Symbology function/services from the "Reference" tab of the "/discovery/symbology/<version>/lookup" API.
    rdp-ref.png

  • @nicola.fiore

    I am so sorry. My previous answer is incorrect.

    Actually, the "rdp.convert_symbols()" function uses the https://api.refinitiv.com/discovery/search/v1/lookup endpoint.

    Therefore, from my test, the limitation of this endpoint is not 3000.

    I got the RealTimeout error when requesting many RICs.

    1660554826397.png

    If you got the same error, you can increase the ReadTimeout value via the configuration, as mentioned in this thread.

    {
      "config-change-notifications-enabled": false,
        "http":{
                "request-timeout": 1000,
                "max-connections": 100,
                "max-keepalive-connections": 20
        } 
    }

    You may need to call this code to load the configurations.

    import refinitiv.dataplatform as rdp
    config = rdp.configure.config
    test_config = rdp.configure.ext_config_mod.config_from_json(
       "c:\\Refinitiv\\RDP.Python\\my-config.json",
       read_from_file=True,
       )
    config.update(test_config)


  • @nicola.fiore

    I got an update from the product team that the lookup endpoint will accept a maximum of 5000 terms (i.e. 5000 symbols to lookup)

  • Hello @nicola.fiore,

    In addition to the helpful information from @Jirapongse and @wasin.w,

    If you are using RD library python to request, and would like to work work with longer instrument lists, you may find it helpful to review this article, with downloadable code to chunk the list prior, to loop through the smaller requests, and to re-assemble the result once all the smaller requests are completed.