Support Board
Date/Time: Fri, 01 Nov 2024 00:12:11 +0000
[Locked] - Bid Ask Depthbars not being calculated
View Count: 3325
| [2019-08-02 14:18:48] |
| Tooth Fairy - Posts: 79 |
|
Can you put an option & a percentage parameter to whether to filter or not? Then everyone is happy?
|
| [2019-08-02 19:34:18] |
| User132748 - Posts: 159 |
|
Please make it 40%. It would cover over 95% of market data so acceptable for now until you find the actual problem and fix it. From my understanding that filtering thing is kinda bypass of not yet identified/found bug so it's a temporary solution and filtering will be totally removed when actual problem found and fixed? We really appreciate your new implementation of depthbars. Looking forward to get an access from ACSIL to new depth data structure. BTW could you confirm that in this new implementation depthbars are being recalculated at EVERY depth change? Ive noticed they are lagging by noticeable amount of time vs DOM depth (read by sc.GetAskMarketDepthEntryAtLevel() function with sum of all available depth levels) which gives much faster depth change info. Same time new depthbars are little faster than T&S total depth records (SC_TS_BIDASKVALUES + TotalBidDepth/TotalAskDepth) which is very nice. Thanks for great work guys! Really appreciated. |
| [2019-08-03 11:44:11] |
| User786248 - Posts: 25 |
|
+1
|
| [2019-08-05 13:12:12] |
| User132748 - Posts: 159 |
|
Unfortunately spikes are back @1968. Offers spike in perfect 20min brackets as in screenshot. Spikes are not visible at T&S depth and also not in previous builds (including 1967) Of course it's better to see those spikes than no bars at all, but we would be grateful if that issue was finally resolved. Thank you! Date Time Of Last Edit: 2019-08-05 13:19:00
|
 spikes1.jpg / V - Attached On 2019-08-05 13:11:13 UTC - Size: 204.74 KB - 560 views |
| [2019-08-05 19:37:42] |
| User132748 - Posts: 159 |
|
Same regular offer spikes on @1969. Second half of trading day on screenshot.
|
| spikes2.jpg / V - Attached On 2019-08-05 19:37:27 UTC - Size: 199.69 KB - 436 views |
| [2019-08-05 20:35:36] |
| User132748 - Posts: 159 |
|
Same thing at my friend's instance. We're both on CQG webapi. @1969 Spikes have excatly same timing but wicks sizes have minor diferences. Date Time Of Last Edit: 2019-08-05 20:44:45
|
| spike3.png / V - Attached On 2019-08-05 20:35:31 UTC - Size: 142.59 KB - 502 views |
| [2019-08-06 02:16:18] |
| |
|
Post # 28, 29 and, 30 only confirm why the filtering is in there to begin with. We had a good reason to do that. We are going to run a test with the Sierra Chart Exchange Data Feed and see if we notice this with the current filtering as it is in 1969. If not, it just has to do with how CQG is providing market depth data with their data feed. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2019-08-06 02:26:00] |
| |
|
So we have started recording market depth data while using the Bid and Ask Depth Bars study, 300 levels of depth, with the SC Exchange Data Feed for CLU19 and we will post what we have tomorrow around this time.
Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2019-08-06 07:23:30] |
| User132748 - Posts: 159 |
|
Spikes are present at 1967 too where imbalance filter is at 80% threshold. Extreme imbalance filtering is what makes this study totally unusable, as we cannot see any bars for hours so it's obviously no way to go. Please find the solution, that allows this study to be used regardless of current bids and offers situation. This study should truely reflect real market image. No data filtering can be justified other than being some work-around of real coding problem. Those spikes are NOT PRESENT on depth being calculated using Time&Sales data structure (I can send my chart and my code). So it's not a problem of data feed as you obviously will see after your recordings. Also they are NOT PRESENT at 1941 build with your own previous implementation using T&S data. Your recent implementation of Depthbars study and depth processing in general is great initiative but I hope you understand it's not acceptable to have no depth bars for hours. And although we value your work regarding your own data feeds and it's possible we would use them at some point, it's hard to accept an explanation that it's some CQG glitch, existing globally in geographical sense and lasting for days without anyone else noticing it despite of the users of Sierra Chart. If you cannot deal with those spikes please REMOVE THE FILTERING COMPLETELY. It's better to see those spikes than no bars at all :-( Thank you Date Time Of Last Edit: 2019-08-06 07:43:23
|
Private File |
| [2019-08-06 08:49:41] |
| User132748 - Posts: 159 |
|
To be clear those "spikes" are in fact massive drops and pick ups of offers, counting 1000 offers on each drop on several levels starting from level 200 to 500. Screenshots below are studies overlayed from subcharts having decreased number of depth levels in Chart Settings. These drops are not visible on any other client using CQG data feed. Date Time Of Last Edit: 2019-08-06 09:17:48
|
| spike5.jpg / V - Attached On 2019-08-06 08:46:30 UTC - Size: 181.62 KB - 516 views offersdrop1.jpg / V - Attached On 2019-08-06 09:10:29 UTC - Size: 153.63 KB - 512 views |
| [2019-08-06 11:59:41] |
| zippyzip - Posts: 293 |
|
no reason to post sierra chart data to us , there are thousands of loyal cqg users and i think its only right that you work on a solution to fix this and not blame cqg or promote your data . we all love the platform and love the work you guys do but we are solid cqg users as its the better data. i run this same info on cqg IC and there are no spikes , no glitches so we would love to have same smooth depth read pre 1958 , im confident this will be fixed , thank you guys
|
| [2019-08-06 17:04:53] |
| |
|
If we have to blame CQG, we are most certainly going to blame CQG. And point out the problem. They have problems. Post #35 is pointless. It would be stupid for us to do otherwise. And this is what we see so far using the Sierra Chart Exchange Data Feed: http://www.sierrachart.com/image.php?Image=1565110923972.png Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2019-08-06 17:07:16
|
| [2019-08-06 18:16:50] |
| User132748 - Posts: 159 |
|
Yes, it would be great if you could point out the problem so it could be sent to CQG to fix the issue. Until then seems the problem is on Sierra side because we all see proper CQG data using other clients. Also: CQG datafeed based charts look good using SierraChart builds<=1941 and when using T&S based depth on current builds (see screenshot). So from the point of view of a simple user the issue is somewhere in the new implementation of Depthbars study/new depth processing. With respect to how cool new implementation is in general. Hope you will find the solution soon. Thank you! Date Time Of Last Edit: 2019-08-06 18:24:10
|
| spike6.jpg / V - Attached On 2019-08-06 18:11:42 UTC - Size: 127.76 KB - 386 views |
| [2019-08-06 18:42:13] |
| |
|
In regards to post 37, at this point we are just gathering information and doing comparisons and gradually understanding the problem. This is an involved process. Also CQG will not change how their data feed behaves. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2019-08-06 19:14:03] |
| User132748 - Posts: 159 |
|
In regards to post 37, at this point we are just gathering information and doing comparisons and gradually understanding the problem. This is an involved process.
Thank you! Fingers crossed! |
| [2019-08-06 19:15:52] |
| |
|
This is what we see with the Sierra Chart/Denali Exchange Data Feed: http://www.sierrachart.com/image.php?Image=1565118788689.png And this is what we see with CQG calculating only 300 levels: http://www.sierrachart.com/image.php?Image=1565118824236.png Also the capabilities of the Sierra Chart data feeds are far beyond that of CQG. CQG does not provide historical data for expired futures contracts. That is provided by our own data feeds. We are adding that in. CQG does not provide that. We also provide 300 levels of market depth across all symbols. And we will take that up to 500. This is also high quality low latency data feed and you can control the number of levels of market depth. For those who require advanced and reliable market depth analysis it is not too much to ask to pay extra for a better data feed and as we continue along our progress of creating unified solutions, we will see about lowering the price. But eventually for those of you who just want CME data when using CQG, there is no reason for us to even be supporting the CQG data feed for CME data and eventually that will just get phased out by us, only our own data feed will be available. But at that time, we will definitely have it more cost competitive. But it also needs to be understood the capabilities of what we offer are far beyond CQG. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2019-08-06 19:18:00
|
| [2019-08-08 10:29:28] |
| |
|
The last two posts have been deleted. There was too much incorrect information, and we cannot take time answering inaccuracies. This hurts everyone. We will continue to delete posts as needed. The following is the logging showing the behavior of the CQG feed at the time of these "spikes": Message Log: RecordMarketDepthChanges | Num bid levels=281. Num ask levels=347 | 2019-08-08 05:50:17.572 Message Log: RecordMarketDepthChanges | Num bid levels=281. Num ask levels=347 | 2019-08-08 05:50:17.694 Message Log: RecordMarketDepthChanges | Num bid levels=281. Num ask levels=347 | 2019-08-08 05:50:18.069 Message Log: RecordMarketDepthChanges | Num bid levels=281. Num ask levels=347 | 2019-08-08 05:50:18.154 Message Log: RecordMarketDepthChanges | Num bid levels=31. Num ask levels=0 | 2019-08-08 05:50:18.176 Message Log: CQG WebAPI | AskDepthDelete = 400 | 2019-08-08 05:50:18.177 Message Log: CQG WebAPI | BidDepthDelete = 291 | 2019-08-08 05:50:18.177 Message Log: RecordMarketDepthChanges | Num bid levels=140. Num ask levels=0 | 2019-08-08 05:50:18.179 Message Log: CQG WebAPI | BidDepthUpdate = 109 | 2019-08-08 05:50:18.179 Message Log: RecordMarketDepthChanges | Num bid levels=248. Num ask levels=0 | 2019-08-08 05:50:18.181 Message Log: CQG WebAPI | BidDepthUpdate = 108 | 2019-08-08 05:50:18.181 Message Log: RecordMarketDepthChanges | Num bid levels=279. Num ask levels=0 | 2019-08-08 05:50:18.182 Message Log: CQG WebAPI | BidDepthUpdate = 31 | 2019-08-08 05:50:18.182 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=66 | 2019-08-08 05:50:18.184 Message Log: CQG WebAPI | AskDepthUpdate = 66 | 2019-08-08 05:50:18.185 Message Log: CQG WebAPI | BidDepthUpdate = 43 | 2019-08-08 05:50:18.185 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=174 | 2019-08-08 05:50:18.186 Message Log: CQG WebAPI | AskDepthUpdate = 108 | 2019-08-08 05:50:18.186 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=283 | 2019-08-08 05:50:18.190 Message Log: CQG WebAPI | AskDepthUpdate = 109 | 2019-08-08 05:50:18.190 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=336 | 2019-08-08 05:50:18.193 Message Log: CQG WebAPI | AskDepthUpdate = 108 | 2019-08-08 05:50:18.193 Message Log: CQG WebAPI | AskDepthUpdate = 9 | 2019-08-08 05:50:18.194 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=336 | 2019-08-08 05:50:18.286 Message Log: RecordMarketDepthChanges | Num bid levels=283. Num ask levels=336 | 2019-08-08 05:50:18.287 The short answer, is that this is not a properly designed and proper functioning data feed which is usable for advanced market depth analysis. They are applying some kind of hack in order to try to maintain an accurate book by rebuilding the book every 20 minutes. This is inefficient. The Sierra Chart data feeds do not do anything like this, because it is unnecessary if a data feed is properly implemented. Additionally what we observe is that CQG is not providing full market depth unlike the Sierra Chart /Denali Exchange Data Feeds. Therefore, the depth data it is providing, for the purpose for which you want to use it for, is definitively not completely accurate. So the number of actual levels available is not being sent by CQG. It is about 90% complete though. Just a rough estimate. The data feed is inefficient and not well designed because every 20 minutes it is doing mass deletions of levels and then restoring those. It is this mass deletion and restoring which is causing the imbalances . These actions are spanning more than one message which is difficult to then determine the right point at which to mark to do the calculations. You need to understand Sierra Chart is performing a very accurate calculations now with Bid and Ask Depth Bars. Furthermore, CQG data is processed on the primary thread rather than a background thread with the Sierra Chart data feeds. We have decades of experience trying to work with services like CQG and it is easier and more effective for us to have direct control over issues like these with our data service. And in this particular case it is only possible to do it right ourselves, not using other feeds. If you are looking for quality you should use the Sierra Chart data feeds. CQG is not meeting a proper standard here. We will see what we can do. But any type of fix for this is going to overall affect other types of calculations like pulling and stacking. This is why we said the CQG data feed for CME data is going to get phased out by us. That is going to happen but we do not know the timing. With the Sierra Chart data feeds, there is a flag variable which was recently added which indicates the proper point in time based on the origin CME multicast feed, to do the calculations. This is high precision. With the behavior of the CQG feed, you have to ask yourself whether pulling and stacking data is even accurate. It is not. It just is not. If you want proper market depth analysis then we have a data feed you can use. So at this point our position is we are going to look at what we can do but it is going to be far from perfect. We will also write up some information just like we do with all of these CQG disconnections and refer users to that. We can not answer all of the CQG nonsense anymore. If you want a quality and properly designed data feed, use this one: Denali Exchange Data Feed We are sick and tired of dealing with problems from others out there. And we are not going to put up with this anymore. This thread is now locked and we are going to finalize our position on CQG regarding this issue. If we lose business, so be it. But we are going to give you something better. And that is available now. We will see what we can do on better pricing. But it is already actually a better price than CQG standard pricing. CQG standard pricing is $25 a month through any other broker when you consider the monthly fees, and also historical data. And this "spikes" problem only applies to the CL when using CQG. It does not apply to other symbols. And when you are using our data feed, you can still use CQG for order routing. You just need to contact your broker, and have them cancel the CME exchange fees so you do not pay those twice. And we are also working with AMP to have this data feed available to be paid through your brokerage account. We hoped it would be this month. But this probably is not going to be ready until September. And we will see about what special pricing we can offer you. But for the time being not likely less than 20 USD a month considering the capabilities which is far beyond what CQG offers. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2019-08-08 10:58:43
|
| [2019-08-08 11:04:06] |
| |
|
To be clear, we are tired of dealing with external service problems. This is why we offer our own quality data feeds. If you want quality analysis of CL futures market depth data, your only choice is to use this data feed: Denali Exchange Data Feed Our position on this is firm. Furthermore, this data feed will be available through your broker AMP if you are with AMP and you can still use CQG for order routing, you do not have to use ours and we will see about what we can do for favorable pricing. Does CQG support numerous external data feeds? Other than from the exchanges, the answer is absolutely not. They have just one CME data feed and that is their own. So we are offering our own as well.. This only makes sense. It makes sense that everyone is getting consistent and reliable and accurate data for Sierra Chart. It is completely nuts to be supporting so many different data feeds, everyone having different problems and for us to be wasting time answering all of these questions. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2019-08-08 11:08:09
|
| [2019-08-08 11:26:20] |
| |
|
And one final piece of information is that we will increase the number of levels of market depth to 500 levels with the Sierra Chart data feeds including the Denali feed. This is not a problem because we have the data. And this is 500 levels across all CME major futures contracts not just the CL.
Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
To post a message in this thread, you need to log in with your Sierra Chart account: