Bad news
Do you have any idea of a release date for 4? Are we talking weeks, months, or unknown??
Geoff
| Dukascopy Support Board http://www.dukascopy.com/swiss/english/forex/jforex/forum/ |
|
| Bad data in Renko feed http://www.dukascopy.com/swiss/english/forex/jforex/forum/viewtopic.php?f=85&t=50994 |
Page 3 of 5 |
| Author: | giorgio58 [ Sun 07 Sep, 2014, 13:44 ] |
| Post subject: | Re: Bad data in Renko feed |
again nothing news? |
|
| Author: | Platform Support [ Fri 12 Sep, 2014, 17:16 ] |
| Post subject: | Re: Bad data in Renko feed |
We have been busy with CFD launching lately. New renko system is coming soon. |
|
| Author: | scotpip [ Wed 17 Sep, 2014, 16:58 ] |
| Post subject: | Re: Bad data in Renko feed |
Hi Guys I realise that you've got a lot on your roadmap right now, but in April you said that we could expect a fix pretty soon, and now it's mid-Sept. It's well over 6 months since this bug was raised. I have a bot design that's working on paper, but I need access to a couple of Indicators in the Stragety. It's pretty easy to re-implement Renko in my own code, but I really would prefer not to have to re-implement the indicators too. So it would be VERY much appreciated if you could get this into JForex 2. Given that you seem to have working code, would this be a big job? |
|
| Author: | Platform Support [ Thu 18 Sep, 2014, 15:21 ] |
| Post subject: | Re: Bad data in Renko feed |
scotpip wrote: Given that you seem to have working code, would this be a big job? It is. |
|
| Author: | scotpip [ Thu 18 Sep, 2014, 15:39 ] |
| Post subject: | Re: Bad data in Renko feed |
Bad news Do you have any idea of a release date for 4? Are we talking weeks, months, or unknown?? Geoff |
|
| Author: | scotpip [ Tue 23 Sep, 2014, 20:51 ] |
| Post subject: | Re: Bad data in Renko feed |
By the way, for anyone following this I've found a half-decent workaround for getting accurate data on new Renko bars without having to analyse every tick to make my own (this could get processor intensive when following lots of instruments in lots of price ranges). 1) Set up a Renko feed. 2) When onFeedData is notified of the new bar, ignore the bad data in the feedData param. 3) In onFeedData, use history.getRenkoBar() to get correct data on the latest bar. IBar bar = history.getRenkoBar(Instrument.EURUSD, OfferSide.BID, PriceRange.ONE_PIP, 0); Execution time for the call to history is acceptable for my purposes, so this should work OK for me. You'd better check this for your own scenario. |
|
| Author: | jamesadk [ Wed 24 Sep, 2014, 13:15 ] |
| Post subject: | Re: Bad data in Renko feed |
Scotpip, You are correct in way, I have been doing the same thing myself but there is a bug with that as well. I do not have my code to hand so I cannot show you my work around but basically there is an error when there is a reversal: E.g. on an UP, UP, DOWN brick formation, you will be able to see if you create some variables for the open, high, low and close prices for all three bricks, that the second UP brick (i.e. the high) will have open price==lowprice==close price where all three are less than high price once the down brick has formed, it seems that they link the close price to the open price of the next brick which is, even for renko, incorrect. It is of course the opposite for a DOWN, DOWN, UP formation. I had to create a work around to properly assign the closing prices of each brick: e.g. if(open price==low price && low price==close price && open price < high price) close price = high price; I'm still working on my own work around for creating an array based on corrected close prices so that I can try and get the indicators to work off the array rather than the price feed. If I get it to work I will post it. Once again I'd like to thank you on voicing your frustration that this matter has been dragging on for so long, it's a feeling I share. It does seem, despite everyone's best efforts in this forum to raise the awareness of this issue, that the team behind the product aren't interested in fixing their mistake with much haste and would apparently rather focus on new releases. I have even called several members at Dukascopy about this and keep getting pointless responses. I sit and wait.... |
|
| Author: | scotpip [ Wed 24 Sep, 2014, 14:16 ] |
| Post subject: | Re: Bad data in Renko feed |
James - thanks for the heads-up. I had only given the history data a quick check and it looked OK. I'm into heavier testing now, so will look at this carefully. |
|
| Author: | scotpip [ Wed 24 Sep, 2014, 14:22 ] |
| Post subject: | Re: Bad data in Renko feed |
I'd be very grateful if you find a way to get the indicators to work. At present I'm working on a strategy that uses a custom Renko based indicator in the background, but that trades off the candle chart so I get working indicators. I'd much rather be working with Renko end-to-end though... I'm so frustrated with this that I've investigated other brokers and platforms, but the Dukas offering is hard to beat. I just wish they'd focus more on getting things to work reliably rather than on producing new shiny things 
|
|
| Author: | jamesadk [ Wed 24 Sep, 2014, 15:01 ] |
| Post subject: | Re: Bad data in Renko feed |
Couldn't agree more. I'd MUCH rather have a platform that works 100% where I can fully rely on the data that is being given to me than being able to choose between a million different colours for my moving averages, but as you said it's hard to beat Duka right now. Just seems that the developers have forgotten that you can't build a mansion on sand foundations. |
|
| Author: | Platform Support [ Fri 26 Sep, 2014, 16:26 ] |
| Post subject: | Re: Bad data in Renko feed |
I can only agree on this. However it is not the shiny look or million of collors which have stolen our time, but adding new instruments. CFD are available on LIVE now and we can get back to were we stopped. |
|
| Author: | jamesadk [ Fri 26 Sep, 2014, 21:33 ] |
| Post subject: | Re: Bad data in Renko feed |
Well that at least sounds promising! Scotpip: As per my previous message about correctly deriving the closing prices of bricks, I have found a more reliable way of doing it since the first way I suggested still on occasions wouldn't work as expected. I changed my technique to compare the high and low price of one brick to the previous brick and therefore derive the prices (since it only ever seemed that the high and low prices were completely reliable, the open and close prices would often change as I previously said). E.g. if(current low==previous high && current high>previous high){ current close=current high current open=current low } else if (current high==previous low && current low<previous low){ current close=current low current open = current high } Hope that helps. |
|
| Author: | scotpip [ Sat 27 Sep, 2014, 16:18 ] |
| Post subject: | Re: Bad data in Renko feed |
Hi James Your code looked familiar, and I realised I'm already doing something similar when I'm building my history cache on launch. It's not code I'd touched for a bit and I'd simply forgotten about the bug! Thanks for the reminder. |
|
| Author: | madbenjies [ Tue 11 Nov, 2014, 19:20 ] |
| Post subject: | Re: Bad data in Renko feed |
Hello! Are there any news regarding renko charts? The open and close prices are still messed up and doesn't make any sense both in real time and historical tester! This is quite an issue in my opinion! |
|
| Author: | Platform Support [ Thu 20 Nov, 2014, 18:03 ] |
| Post subject: | Re: Bad data in Renko feed |
We plan to release the new version next week. |
|
| Author: | scotpip [ Tue 25 Nov, 2014, 12:16 ] | ||
| Post subject: | Re: Bad data in Renko feed | ||
Hi Great that this issue is finally being addressed! I'll be checking the feed later today, but it's immediately obvious that there's a pretty major bug with the charting - it simply isn't updating as the price moves into the zone that should trigger a new bar. I've enclosed a screenshot that illustrates the issue.
|
|||
| Author: | Platform Support [ Wed 26 Nov, 2014, 12:13 ] |
| Post subject: | Re: Bad data in Renko feed |
Version 2.42.2 is out on DEMO. More info on Renko is available here: https://www.dukascopy.com/wiki/#Charts_widget under Renko tab. |
|
| Author: | scotpip [ Wed 26 Nov, 2014, 13:38 ] |
| Post subject: | Re: Bad data in Renko feed |
Thanks for the update. Great that we are finally close to a working Renko implementation! The new system looks pretty flexible and a quick check suggests that it is working OK, apart from the Renko wicks issue which you are already aware of. But I feel that you might want to consider the defaults, as I don't think I'll be the only user who finds them confusing. To have a very fast default Renko period of 1 pip and a very slow Session Length of 1 minute leads to odd looking behavior in a fast-moving chart, with large price movements being ignored. When I saw this, I simply assumed it was a bug. My suggestion would be that you either increase the default period or reduce the default Session Length so they are in better balance. |
|
| Author: | jamesadk [ Wed 26 Nov, 2014, 14:13 ] |
| Post subject: | Re: Bad data in Renko feed |
I'm not sure what has been changed in the way the graphs are calculated but this latest release no longer matches up against my previous copies of charts that I make every day for historical reference (in terms of the actual brick pattern). Is it defaulting to calculating bricks on 1 minute data as opposed to ticks? I have looked for the preferences page as shown in the link above from API support but I can't find it anywhere in the latest build. It seems to be a screen shot from the settings page of JForex4 not JForex2.. |
|
| Author: | scotpip [ Wed 26 Nov, 2014, 14:33 ] |
| Post subject: | Re: Bad data in Renko feed |
Just checked the Renko wick feature on the charts, and it seems to be bugged. For long bars, you can only get a wick on the short (open) end, and vice-versa for short bars. So the wicks are running in the wrong direction! Here is a shot of the current JForex demo release: And here is a shot of a correct implementation from Sierra: |
|
| Page 3 of 5 |
