For a deeper look into our Eikon Data API, look into:

Overview |  Quickstart |  Documentation |  Downloads |  Tutorials |  Articles

question

Upvotes
Accepted
646 10 18 26

Json structure documentation

Hello,

I can see on this forum that in multiply answers there is a reference to raw json request/response objects. Also python get_timeseries and get_data has 'raw_output' parameter that can be set to 1 and the data will come in json format.

We have some concerns about the usage of raw json data in production environment.

Is there any documentation about the format of json requests/responses for different type of data requests?

And the more important question, can we assume that the structure will not be silently changed without any notifications to the users?

Thanks

eikoneikon-data-apipythonworkspaceworkspace-data-apirefinitiv-dataplatform-eikonjson
icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

1 Answer

Upvotes
Accepted
39.3k 76 11 27

The schema for JSON responses is purposely not publicly documented yet, as we do expect changes. The raw_output parameter was introduced primarily to aid troubleshooting, not to encourage the use of data in raw JSON format.
This said Eikon Data APIs for Python library retrieves JSON from the Web service and returns pandas dataframe (by default). This fully supported library must interpret JSON documents returned by the Web service. Any breaking changes to the structure of the JSON returned by the Web service would require changes to the client side Python library, which means they cannot be introduced silently.
We do expect that the structure of the JSON may change as we introduce new data distribution platform. It's very unlikely that this new platform will ever be used to feed existing version of Eikon (Eikon 4). It's very likely that this platform will only deliver data to the new version of Eikon known as Eikon 5, which is currently expected to see the first release mid next year.

icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

@Alex Putkov.

Will Python Data API functions get_timeseries and get_data still work in the same way in terms of arguments and response view with Eikon 5?

As I mentioned in the question, I saw a few times on this forum, that for some cases you suggest to use send_json_request which is the way to use raw json. So I thought you're fine if the clients us it..

We only recommend using send_json_request when there's no alternative at the time. The design of the Python library for the new platform has not been finalized yet. At this point it seems likely that there will be some breaking changes aimed primarily to improve the usability. But even with some breaking changes I anticipate that the migration should be very straightforward. If you're using send_json_request the migration is likely to be less straightforward.

Understood. Thank you

Regarding raw_output, I can see a problem with DataFrame structure, it's not possible to use multiply tickers in time series request. However, it works fine with raw_output=1. Which means this limitation is not from web-server, but because of DataFrame datatype itself.