Join the 80,000 other DTN customers who enjoy the fastest, most reliable data available. There is no better value than DTN!

(Move your cursor to this area to pause scrolling)




"I'm satisfied with IQFeed. It's the most reliable and fastest quote feed I have ever used. Although I'm a resident in China, it's still very fast!" - Comment from Xiaofei
"After all the anxiety I had with my previous data provider it is a relief not to have to worry about data speed and integrity." - Comment from Eamonn
"I was on the phone with a friend who uses CQG and right after the Fed announcement, CQG was as much as 30 seconds behind DTN.IQ. Some quotes were off by as much as 15-18 cents. Your feed never missed a beat." - Comment from Roger
"Boy, probably spent a thousand hours trying to get ******* API to work right. And now two hours to have something running with IQFeed. Hmmm, guess I was pretty stupid to fight rather than switch all this time. And have gotten more customer service from you guys already than total from them… in five years." - Comment from Jim
"My broker in Davenport suggested I give you a try as he uses your service and says its the best." - Comment from Bill via RT Chat
"DTN has never given me problems. It is incredibly stable. In fact I've occasionally lost the data feed from Interactive Brokers, but still been able to trade because I'm getting good data from DTN." - Comment from Leighton
"I ran your IQFeed DDE vs. my broker vs. a level II window for some slow-moving options. I would see the level II quote change, then your feed update instantaneously. My broker's DDE, however, would take as much as 30 seconds to update. I am not chasing milliseconds, but half a minute is unacceptable." - Comment from Rob
"I am very happy I changed. I love the product, but more so I am thrilled with Tech Support. You are knowledgeable, polite, pleasant and professional." - Comment from Pat
"Very impressed with the quality of your feed - ******* is a real donkey in comparison." - Comment from A.C. via Email
"You are much better than lawyers or the phone company because you answer the phone when I call! I just love your customer service." - Comment from Isreal
Home  Search  Register  Login  Recent Posts

Information on DTN's Industries:
DTN Oil & Gas | DTN Trading | DTN Agriculture | DTN Weather
Follow DTNMarkets on Twitter
DTN.IQ/IQFeed on Twitter
DTN News and Analysis on Twitter
»Forums Index »Archive (2017 and earlier) »IQFeed Developer Support »RequestMinuteHistory is calendar days, RequestDayHistory is trading days
Author Topic: RequestMinuteHistory is calendar days, RequestDayHistory is trading days (4 messages, Page 1 of 1)

dkobelt
-Interested User-
Posts: 14
Joined: Jan 2, 2007

Can I have More Data please? :)


Posted: Jan 15, 2007 01:47 PM          Msg. 1 of 4
Making a COM call to RequestDayHistory has a parameter for # of days.
This always returns 3 trading days when 3 are asked for.

However, RequestMINUTEHistory has a parameter for # of days, but the results depend on the day of the week, or holidays.
A weekend and a market holiday (or other special observances) means I might not get any days, or 1 or 2 or 3 days of data.

Can the Minute history be made consistent like Day, Week and Month history requests?

I know you have the data to figure this out :)

I can add code on my side, but you did such good job with Day/Week/Month. Also, its not simple to recongize imprompto market closings for presidents, etc. Whereas you have the trading days known already.

Thank you

DTN_Steve_S
-DTN Guru-
Posts: 2093
Joined: Nov 21, 2005


Posted: Jan 15, 2007 03:37 PM          Msg. 2 of 4
There is actually nothing built into our system that recognises "trading days".

The reasoning this is done this way is so that it is universal to all markets and exchanges.

All of our history data is built from the trades that come through the Level 1 Feed.

Our servers have a nightly process that builds daily/weekly/monthly data from the current day's ticks. If there are no ticks, then there is no daily datapoint created. If there is even a single trade, then a daily datapoint gets built.

This same behavior also extends to minute history vs tick history. If no trades happen in an interval, no minute datapoint is returned for that interval.

While it would be possible (although not necessarily "easy" due to the way the system is setup) to account for non-traded days in minute history, it would require a change in protocol that would require most of our 3rd Party Applications to update thier software to accomodate.

As a result, it is not likely that a change such as this will happen soon but it is certainly a possibility for the future at some point.

FullyArticulate
-DTN Guru-
Posts: 332
Joined: Sep 22, 2005


Posted: Jan 16, 2007 11:05 AM          Msg. 3 of 4
Hi Steve,

I think he's actually arguing the opposite. Today, a HD query gives you trading days, whereas an HM query gives you calendar days.

If it's a Sunday, and you query one day of daily history, you'll get Friday's data. If you query one day of minute history, you get nothing back at all.

While I agree there'd be a 3rd party application problem, this inconsistency is confusing and sometimes difficult to work around. Somewhere in your server, you must have a comparison to calendar date when an HM query is done, but a simple "days of datapoints" comparison when an HD query is done.

DTN_Steve_S
-DTN Guru-
Posts: 2093
Joined: Nov 21, 2005


Posted: Jan 16, 2007 12:24 PM          Msg. 4 of 4
You are correct that there must be a calendar date calculation somewhere in the history server but I would guess (I don't deal with the server code at all) that there is not a "days of data points" comparison done.

The reason, is that with a daily/weekly/monthly/"lastX ticks" request, you are requesting a specific number of data points. Whereas, with a minute/"tick days" request, you are requesting an indeterminate number of data points so the date calulation has to be made. Due to the way the history servers work (as described above), they build trading days automaticly for requests that you specify the number of data points.

We understand that there are flaws with the current system and it should be possible to make this an option in the future. Right now it wouldn't be an easy thing to implement but we will keep it in mind for future development.
 

 

Time: Mon May 20, 2024 1:31 AM CFBB v1.2.0 11 ms.
© AderSoftware 2002-2003