Login Page - Create Account

Support Board


Date/Time: Thu, 31 Oct 2024 23:49:33 +0000



Orderflowanalytics-style reversal bars.

View Count: 31865

[2015-08-26 01:03:24]
ejtrader - Posts: 688
Thanks Norvik. This cpp file works fine as expected :)
[2015-08-26 14:00:46]
ejtrader - Posts: 688
Just that doesn't seem to update the bars in realtime though. If I refresh the chart it displays the latest. Do your chart reflect the realtime prices (compared to regular chart).

screencast.com/t/sxmiN7d4zx4o
Date Time Of Last Edit: 2015-08-26 14:01:07
[2015-08-26 15:26:40]
crazybears - Posts: 314
Hi guys

i try to do something similar .

e.g the size of reversal bar fixed at 1 point (10 ticks) and a new bar with a different fixed size,e.g 5 ticks
since code it's not optimized it requires great resources , these days using it requires more than 2gb of memory (CL TF YM NQ ES).
if someone is interested and want try to optimize it , i can share the code , would be great if could be add to SC renko bar types.
Date Time Of Last Edit: 2015-08-26 15:27:18
[2015-08-26 16:37:48]
ejtrader - Posts: 688
crazybears - if it is functional but just has the performance issues - please post the cpp & chartbook. would take a look at it :)

thanks
[2015-08-26 17:32:15]
crazybears - Posts: 314
Hi Ejtrader

thanks .
would be great if someone could help with NinjaTrader code .
I would translate the indicator you find in this link for Sierra chart.
I need to understand more in depth what is the logic behind it .

https://www.bigmiketrading.com/elite-circle/5995-using-llma-low-lag-moving-average-34.html#post514892
Attachment Deleted.
attachmentBetterRenko2.cpp - Attached On 2015-08-26 17:31:32 UTC - Size: 12.91 KB - 660 views
[2015-08-26 18:23:02]
ejtrader - Posts: 688
crazybears - will comment about the question in BMT forum.

Thanks
[2015-08-27 10:43:03]
Zdislav - Posts: 73
Norvik:
Thank you very much for the code. I put the same study together on my own, but I had some performance issues... I just realized that I didn't optimized the code during compilation (in VS2012). Compiling through SC compiler makes the DLL files much smaller and the code runs quicker. I have one recomendation to improve your logic: When comparing "truerange" value against "targetvalue" use GREATER_EQUAL_OPERATOR in the function instead of GREATER_OPERATOR. This way the final chart looks exactly the same as the original OFA chart (when comparing the look from Youtube daily videos).

ejtrader: The underlying chart must be set to "1 tick chart". Otherwise the chart is useless giving false information. The study updates correctly in realtime.
[2015-08-27 13:09:29]
ejtrader - Posts: 688
Zdislav - Thanks for the input. Appreciate it.
[2015-08-27 21:50:38]
norvik_ - Posts: 106
Zdislav, thank you for your comment.
As for Visual Studio, try to play with project properties. There are many ways reduce dll size and make it very compact.
[2015-08-28 05:32:42]
Zdislav - Posts: 73
Thank you, guys. You did a good job on this! Have you managed the other aspects of OFA trading method? Volume cluster analysis, VWAP in drawn VP, momentum/exhaustion prints... So far I have manged all of these somehow, but it has incredibly long loading times. Now I tried some optimization parameters as you suggested and it helped a bit. I am able to come up with algorithms and make it work in SC, but the computing performance is my weak point. I am no programmer, but I'm learning. Still quite a lot to improve, but it's "tradable". :-)
[2015-08-28 08:27:55]
crazybears - Posts: 314
Hi Zdislav

do you have some papers about this kond analysis ?
maybe we can try to add them together

thaks

Max
[2015-08-28 11:27:19]
Zdislav - Posts: 73
Hi Max

Well, I think it is worth to try to join our efforts :-) Now I am just pulling thoughts out of my head, but you have a good point to have them on paper. I'll put something together over the weekend.
[2015-08-28 18:44:53]
Zdislav - Posts: 73
So here are my ideas about momentum mapping study. I managed to make the simple one and I plan to make the real OFA mapping. I hope you can understand what I mean... If you find something wrong, please let me know.

Thanks

Petr

PS: Below you can see how it looks like... Trading Gold futures according to OFA in Sierra Chart with 1 contract on SIM.
(for explanation: mapping are the blue triangles, green-clusters of buying effort, brown-clusters of selling effort, yellowish neutral volume clusters.)
attachmentOFA Momentum.pdf - Attached On 2015-08-28 18:10:21 UTC - Size: 622.33 KB - 1409 views
image2015-08-28_GC.png / V - Attached On 2015-08-28 18:36:59 UTC - Size: 85.78 KB - 2302 views
[2015-08-30 12:23:17]
crazybears - Posts: 314
ok i have to understand how to do diagonal comparison between ask-bid
[2015-08-30 13:27:52]
Zdislav - Posts: 73
That is simple... when iterating through price levels in each bar,I save the BID price to compare it with ASK price in next iteration (for bullish momentum from low to high iteration). Retrospectively for bearish momentum.
[2015-08-30 13:31:07]
crazybears - Posts: 314
can you post he code
thanks
[2015-08-30 13:39:03]
Zdislav - Posts: 73
Not at the moment,but later today perhaps.
[2015-08-30 20:44:14]
Zdislav - Posts: 73
Here is the code as promissed for OFA Momentum in the simple version according to PDF notes higher...
attachmentZdislav_OFA_Momentum.cpp - Attached On 2015-08-30 20:43:53 UTC - Size: 7.28 KB - 1225 views
[2015-08-31 16:27:12]
ejtrader - Posts: 688
Petr - thanks for sharing this code. It was very clever on your part to mark the momentum triangles :)

Question on Volume Cluster portion (which doesn't seem to be included in the above study) - Do you just pick the highest 3 consecutive prices which has the cumulative volume to be highest or some other logic do you use? What is the criteria you are using to mark the volume cluster on the candle?

thanks
[2015-08-31 17:40:30]
Zdislav - Posts: 73
ejtrader: Yes, that is exactly what I am doing. Take the highest cumulative volume of a given (from some input value) number of consecutive levels. In my case I use 3 levels. It is not exactly the OFA method, but it is close enough and it works pretty well. I haven't tried all the markets, only ES, GC and CL. OFA method seems to use different logic giving more clusters in one candle some times. The color distinction is simply comparing BID and ASK cumulative volumes in that given volume cluster against STRENGTH constant.

e.g.
IF BIDvolume >= 1.618*ASKvolume THEN Cluster color is green, resp. brown. If it does not fit either condition, than the auction is neutral (gold).

PS: My code is not ready for sharing yet. I still need to improve array handling.
[2015-08-31 18:00:45]
ejtrader - Posts: 688
Petr - I think it's pretty close to OFA method (if not exact) :)

Thanks
[2015-08-31 18:07:46]
ejtrader - Posts: 688
crazybears - related to post# 29 - I believe this has to do with using 1 tick base bars.

Please see the post below by SC team.

Modified Renko bar | Post: 67134

Thanks
[2015-08-31 21:39:05]
norvik_ - Posts: 106
ejtrader, sometimes support are wrong.
SC have three examples of customchart realization inside ACS_Source folder, all works fine and based on tick by tick data. Of course, internal SC code of Renko bars is more effective than custom dll, but ACSIL interface for custom bars creation is enough efficient in general
[2015-09-04 13:31:19]
crazybears - Posts: 314
Hi Zdislav

sorry for the late response but i'm busy at work .
looks like ofa momentum repaints .
[2015-09-04 13:46:54]
ejtrader - Posts: 688
crazybears - if you look at the code - it's performing the calculations in realtime.

Any indicator which performs intrabar calculations the behavior would be same I would imagine :) As long as it doesn't repaint "prior" bars - it shouldn't be put in repaint category :)
Date Time Of Last Edit: 2015-09-04 14:02:07
[2015-09-04 13:51:17]
crazybears - Posts: 314
Hi ejtrader

i don't think so , i can't try now but i think it's related to the code .

To post a message in this thread, you need to log in with your Sierra Chart account:

Login

Login Page - Create Account