![]() |
|
|||||||
| Connecting Support for establishing connections to your broker or market data service provider. |
![]() |
|
|
Thread Tools | Display Modes |
|
|
#1 |
|
Certified NinjaScript Consultant
|
I am trying to compare performance of my existing IB data feed relative to a trial on Zen-Fire on the Ninja DOM. Is there any way to have two concurrent DOMs operating on the same computer with different data feeds to enable realtime side by side comparison? I have tried to load both NT V5 and V6 concurrently but received an error. If not, any recommendations on how best compare relative performance?
I would also welcome any input from users that have made the switch from IB to Zen-Fire enabled broker. Thanks. Regards, Whitmark |
|
|
|
|
|
#2 |
|
Administrator
Join Date: Nov 2004
Location: Denver, CO, USA
Posts: 11,163
Thanks: 6
Thanked 45 times in 32 posts
|
You can not have one instance of NT using two different real-time data feeds. NT will always use the firstconnection's data.
Try starting NT6 first, then start NT5, I suspect this should be possible. Connect NT5 to IB, NT6 to Zen and put them side by side.
Ray
NinjaTrader Customer Service |
|
|
|
|
|
#3 |
|
Certified NinjaScript Consultant
|
Thanks, I was able to get V6 and V5 working concurrently by opening eachin the order you specified. In comparing the two DOMs side by side, I did notice:
1)A clear and consistent subsecond advantage of one over the other in terms of speed of DOM last price and contract readout 2) depth of market contracts on the DOM for one datafeed tended to move in jumps vs the other that counted up or down very rapidly 3) nearly instantanious market order fills using theSim101 account without the orangepending confirmation flash up on one datafeed where the otherhad adetectiblepause with theorange pending confirmation flash prior to a fill. 4) Interestingly, on numerous tests, even though I placed the same limit order on NT5 first and then NT6 second (because I can't simultaneously click on both DOMs at the same time), the NT6 was filled a second or moreahead of NT5 with the same simulator settings. While these differences might be telling and only useful for scalping applications, its unclear how much of the improvement I might attribute to the differences in V5 vs V6 platforms vs datafeed differences.Can you shed any light on what differences I can expect between the two Ninja versions that would influenced my observations noted above? Regards, Whitmark |
|
|
|
|
|
#4 |
|
Administrator
Join Date: Nov 2004
Location: Denver, CO, USA
Posts: 11,163
Thanks: 6
Thanked 45 times in 32 posts
|
Items 1 and 2 are data feed related. Item 3 is platform related, NT6 fills market orders immediately and most if not all NT5 versions fills on the next incoming tick which results in the delay you see. Item 4 is data feed related. With Zen-Fire you are getting more ticks than IB and thus the fill algorithm is called more often and couldlikely print morevolume than IB.
You will notice that tick charts with Zen-Fire will look different that IB since you get a lot more data, this is also true for Trading Technologies.
Ray
NinjaTrader Customer Service |
|
|
|
|
|
#5 |
|
Senior Member
Join Date: Feb 2006
Location: Bellevue, Washington, USA
Posts: 190
Thanks: 2
Thanked 2 times in 2 posts
|
Are you able to comment on the amount of data recieved between Tradestation, Trading Technologies and Open Tick?
|
|
|
|
|
|
#6 |
|
Administrator
Join Date: Mar 2005
Location: Bamberg, Germany
Posts: 9,994
Thanks: 0
Thanked 6 times in 6 posts
|
Sorry we are not, but may be the community can.
Dierk
NinjaTrader Customer Service |
|
|
|
|
|
#7 |
|
NinjaTrader Product Manager
Join Date: May 2007
Location: Denver, CO
Posts: 17,458
Thanks: 1
Thanked 106 times in 70 posts
|
From what I can tell OpenTick's data is somewhat on the lacking side. This is to be expected though because it is free so not complaining.
|
|
|
|
![]() |
| Thread Tools | |
| Display Modes | |
|
|