Data Integrity - Splits, gaps, spikes, whatnot.

[ckelly]

Jay,
Just to be sure, I did a reboot of Beta 53 and many of the discrepancies were corrected; however, not all. I still see Volume differences between 5.1 and Beta 53 and the total line for RIMM is way off. This pic at 22:10 tonight and 5.1 still on miami-14.

fyi,
Carol

[ckelly]

Larry,
Maybe take another look - I note differences in Volume in your pic. Additionally, yes, I noted in my post that IB continues to update prices and volume after the market closes.

I knew we were rebooting after midnight - somehow, I guess I missed the memo to reboot AFTER the market opens if we want reliable data on Beta. I'll do that tomorrow and see how it goes.

Carol

[Larry Marchman]

Originally posted by ckelly
Larry,
Maybe take another look - I note differences in Volume in your pic. Additionally, yes, I noted in my post that IB continues to update prices and volume after the market closes.

I knew we were rebooting after midnight - somehow, I guess I missed the memo to reboot AFTER the market opens if we want reliable data on Beta. I'll do that tomorrow and see how it goes.

Carol

Yes, there are volume differences, minor differences, not differences that matter, that represent a problem. This is normal.

What you're seeing with RIMM, I don't see. You are getting a bleed of numbers from the symbol VIP right above it. This used to be a problem in previous builds, but I haven't seen it in a long time. No idea why you're getting it and I'm not. Do you perhaps have those syms duplicated elsewhere in the quotesheet? That used to be the key for one row bleeding it's values into the next row.

LAM

[ckelly]

Larry,
I guess I just don't understand. It seems to me that if two computers are counting, they would come up with the same volume to report - my eyes see significant differences in the numbers and so it seems like one of them isn't working right. But, if you consider it normal, I'll adjust my thinking. As far a making trade decisions, I probably wouldn't change my mind over half a million or so - but, seeing these numbers as significantly different leads me to always ask the question: "What else is not right?" And that leads me to feel very insecure with the entire package. That's just the way I was trained to troubleshoot - others may not look at it that way or even notice.

Additionally, if I understand Jay's explaination, 5.1 is 'deleting' some ticks in order to manage the bandwidth issue. That being the case, I would expect 5.1 volume to be lighter than Beta's. As you can see in my pic, this is not always the case.

As for my quote sheet - what you see is all that is in it - no duplications. So, if RIMM is bleeding over from VIP, it may be only because it's next in the quote sheet.

fyi,
Carol

[Larry Marchman]

Carol,

I'm referring to the volume in your most recent screenshot. Not what you were seeing earlier today. The values for QCharts are artificially manipulated as Jay mentioned with volume adjustments periodically. That QCharts can show a value higher than eSignal means QCharts has been adjusted to the actual market volume, where I suspect eSignal is showing it's actual unadjusted volume.

The real issue is tick volume, whether or not a diminished flow of all possible ticks is a show stopper. Maybe, maybe not. Depends on whether or not one is getting a price match moment to moment when comparing one feed to another. What if, for example, rather than receiving 10 ticks per second for a symbol, one is receiving 5 ticks per second, with the ticks on time, with the price the same.

I run a QCharts T&S physically adjacent to my broker T&S and Level-II. Yes, I see fewer ticks sometimes in QCharts, but nevertheless, the price is tracking, the chart's reflecting the current bid/ask/trade prices. Doesn't affect my ability to hit a limit trade, and market trades fill at expected and reasonable prices. If the ticks lag in price due to throttling, then yes Houston, that's a problem. Different stocks will give different experiences. The higher volume stocks that tick multiple times/sec tend, for me, to not pose a problem if I'm only getting 5 ticks/sec rather than 10, just to use an arbitrary example.

Consider sampling as the context here, like ripping MP3s from CDs. Is there all that much difference in a tune ripped at 160kbps vs 128kbps? Can ya still boogie to either?

LAM

[tradertom]

Larry,

The mp3 analogy is a good one. I can hear a significant difference between 128 and 160 kbps. In fact, I rip CDs to 192 kbps to avoid most of the artifacts that lower bitrates insert into the audio. Use headphones; you will hear what I mean. Sure, one can boogie to lower quality music, but I ain't always listening to boogie.

That said, I would accept 128 bitrate mp3 songs if they were provided at no cost; just to hear the tune but not to really listen to it. If I am paying $164.95 per month, I would expect higher quality sound than 128kbps.

I don't get paid for listening to music though; I do get paid from the markets I follow using QCharts. And when one pays for a service, one expects the service to be reasonably accurate. It seems redundant to get another data provider just to check on the accuracy of QCharts data...

Bottom line for me: these data problems do not instill confidence in the highly touted eSignal feed that is intended to improve QCharts.

Thanks,
Tom

[JayF]

Originally posted by tradertom
Bottom line for me: these data problems do not instill confidence in the highly touted eSignal feed that is intended to improve QCharts.

Tom,

Thanks for the comments. Keep in mind that QCharts 6 is still in Beta and has bugs. The problems that Carol and others have been reporting with the Beta are issues in the QCharts software, and aren't from the eSignal network.

We'll get these things in the beta cleared up in due time, and we'll get to a point where QCharts 6.0 will be more reliable than 5.1.

[ckelly]

Jay,
Continuing our comparisons on Volume/Last/Net from yesterday, this pic AFTER a Beta 53 reboot. Still see differences in prices and volume between the two.

Additionally (although not shown), this morning, PBR charts were different from my broker by 0.75. Unfortunately, by the time I realized this and the charts suddenly jumped to catch up - it cost me. Drat.

fyi,
Carol

[JayF]

As I posted yesterday, we have a known problem in QCharts 5.1 with the Volume Only messages not catching the volume up like they should be, and we're in the process of investigating this. We aren't in need of any further reports of volume issues. That being said, I appreciate the efforts put forth on the comparisons.

As an aside, let's try to keep the QCharts 6.0 Beta reports to the QCharts Beta Forum, and Data Integrity reports on QCharts 5.1 on this thread.

[ckelly]

Jay,
This in 5.1.0.21 - I have only seen this weird occurance on the symbols T (multiple) & SHLD (once). It occurs on M/W/D charts at random. Note all the indicators appear correct, however, the moving averages and toggle auto wave jump way out of the BB's.

Yes, I have rebuilt this chart multiple times.

fyi,
Carol

[ckelly]

NKE split this morning - the intraday's show the adjustment, however, the D/W/M charts do not. Please correct.

Thanks,
Carol

[JayF]

The back-end process that automatically updates CorpActs stopped functioning and our engineers are working on correcting that ASAP.

[bobscott]

Jay,

I looked at both the CorpActs.INF and UNF just after Carol's post (pre-market open) and both contained the NKE split but the daily QC6-53 plot was screwed up. I re-started QC6-53 after market open and the the daily plot was corrected. Maybe this was just another reset funny.

~Bob

[ckelly]

May have been a reset funny on QC6-53, I don't know. For me, both 5.1.0.21 and Beta were off on NKE until mid-morning and then they suddenly jumped to correct.

fyi,
Carol

[emboehm]

Missing intraday data for NYSE:AEO

I'm missing intraday data (233 min chart, specifically) for NYSE:AEO. I see the data on the D and 55, but nothing after 09:30 3/19. The data disappeared a couple of days ago