Difference between TR.PriceHigh, TR.Volume from Dex2 and High, Volume from Data API time-series

igorg

Is there any difference between the data from :

TR.Volume (COM API Dex2) (with SDATE parameter in the past) and VOLUME from time-series request for the same day (Interval: 'daily') (Data API) ?

Zhenya Kovalyov

@igorg this is a content question, I suggest you contact Refinitv support desk for that.

Dex2 is a way to access the reference database, while RHistory or the .NET Time series API - the time series database. So, your question really is about the difference of content between two databases: time series delivered over reference database (TR.Volume) and time series database ('.Volume' or 'TRDPRC_1.Volume).

I had a quick look, I have not spotted any differences other then real-time blending of today's value into the time series request:

I would guess that there might be a difference in the scaling factor for high volumes, but the content specialist from the support desk should confirm.

igorg

@Zhenya Kovalyov,

I expected this answer and did contact them first.

They suggested to contact you because it's related to API requests.

Let me clarify my question.

Sometimes we need to get the data from the past. Eikon Data API (python) allows to request only for Volume, Open, Close, High, Low fields. But not VWAP.

There are 2 ways to get these fields:

1. Com API dex2 request (TR.PriceOpen, TR.PriceHigh...)

2. Data API time-series request. But VWAP is not available.

if we can assume that the data from 2 sources is the same, I can run only 1 request using dex2. If not, I have to run time-series request and then separate request using Com API for VWAP field.

Zhenya Kovalyov

https://community.developers.refinitiv.com/discussion/comment/35992#Comment_35992

@igorg I would think that a single Dex2 request should suffice, however, I have requested additional info on this and will let you know should there be any exceptions.

igorg

https://community.developers.refinitiv.com/discussion/comment/35995#Comment_35995

Thank you. I appreciate it.

igorg

The follow up question. Which type of request will work faster? Dex2 or time-series 'daily' for 1 day in the past for the same fields.

----------------------------------------------------------------

I just tested dex2 for

1. around 2000 tickers in 1 request

2. fields: TR.VOLUME, TR.OPENPRICE, TR.CLOSEPRICE, TR.HIGHPRICE, TR.LOWPRICE and TR.TSVWAP

3. SDATE: 2018-11-15

It takes about 4 minutes to get a response. Is there some explanation why this request takes so long time? I ran many different requests and this one is the slowest. For example, similar amount of data in rtx snapshot request take 5-10 times less time.

Zhenya Kovalyov

https://community.developers.refinitiv.com/discussion/comment/36085#Comment_36085

@igorg you can not compare the two, RTX uses a completely different infrastructure and data sources. It is slower, but 4 minutes is too long. Could you benchmark a request with 10 and 100 instruments respectively. Does it still take the same amount of time?

igorg

https://community.developers.refinitiv.com/discussion/comment/36087#Comment_36087

ok,

1950 instruments within 1 request : about 4 minutes

195 requests with 10 instruments in each : 7 minutes. So about 2 seconds per a request

20 requests with 100 instruments in each: 3:45. So about 11 seconds per a request.

I restarted Eikon before each test in case it caches some data in the memory.

Zhenya Kovalyov

https://community.developers.refinitiv.com/discussion/comment/36090#Comment_36090

@igorg i will have a look at what can be done and let you know.

igorg

https://community.developers.refinitiv.com/discussion/comment/36090#Comment_36090

2. fields: TR.VOLUME, TR.OPENPRICE, TR.CLOSEPRICE, TR.HIGHPRICE, TR.LOWPRICE and TR.TSVWAP

3. SDATE: 2018-11-15

igorg

https://community.developers.refinitiv.com/discussion/comment/36103#Comment_36103

Also, this type of request returns the error "Server response is not valid" in about 15% of all requests. I changed 2000 to 500 instruments per a request. No improvements. And it's still slow.

Zhenya Kovalyov

@Jirapongse not really, i need to chase.