a trading strategy based on Elliott Wave Theory - page 48

New comment
 
Made a channel drawing script with price type selection. The difference turned out to be not so big and the method by Clos is more conservative.

This is channel by (High[]+Low[])/2
2006.06.13 22:10:45 PM ChannelScrypt USDCHF,H1: Script<br / translate="no"> Hurst=0.6397 Hurst(VG)=0.4546 channel length 77 bars, 3 sigma width=1003 points
2006.06.13 22:10:45 ChannelScrypt USDCHF,H1: found 576 channels that meet the criterion of 1000 bars
2006.06.13 22:10:45 PM ChannelScrypt USDCHF,H1: They are in 3 series
2006.06.13 22:10:43 ChannelScrypt USDCHF,H1: initialized
2006.06.13 22:10:35 ChannelScrypt USDCHF,H1: loaded successfully


This is the channel by Close[]
2006.06.13 22:10:33 ChannelScrypt USDCHF,H1: removed
2006.06.13 22:10:33 ChannelScrypt USDCHF,H1: deinitialized
2006.06.13 22:10:33 ChannelScrypt USDCHF,H1: Execute deinit()
2006.06.13 22:10:17 ChannelScrypt USDCHF,H1: Script
Hurst=0.5113 Hurst(VG)=0.3675 channel length 76 bars, width 3 sigmas=1003 pips
2006.06.13 22:10:17 ChannelScrypt USDCHF,H1: Found 578 channels satisfying criterion on 1000 bars
2006.06.13 22:10:17 17 ChannelScrypt USDCHF,H1: They are in 3 series
2006.06.13 22:10:16 ChannelScrypt USDCHF,H1: initialized
 
Respect to all who told and asked!

Somewhere on page 19-20 Solandr, posted a picture to reconcile the channels found. I think it's not a bad idea to check if you understand the criteria for channel selection, for those who are new to the system.
Can someone try to find the channels that were shown in the picture Vladislava, I have using the criterion CKO2/3>CKO and the value of the confidence interval of 2.58 *COC (99%), these channels are not found as well, and if I take as parameters 3.29 *COC (99.9%) and just choose the lowest RMS, it turns out very similar
, so I have doubts about how I select, I would be grateful if someone who has implemented the search channels show a picture at this time. Here is my picture also with the "wrong" set of criteria
 
Jhonny, the picture was posted by Vladislav, as far as I remember. It's a bit simpler there - I identified the following levels:
60% (0.84 sigma)
1 sigma
90% (1.6 sigma)
2 sigma
3 sigma
and one unidentified (I suspect 80%).
But the exact percentage is not so important, in principle.

ZS At what prices did you draw the channels ?
 
Maybe somebody tried to find the channels that were shown in Vladislava's picture, when I use CKO2/3>CKO criterion and confidence interval value 2.58*SCO(99%) these channels are not found in any way and if I take 3.29*SCO(99.9%) for parameters and just choose the one with the lowest RMS then I get very similar results

I also using 99% probability I did not get exactly such channels as Vladislava did (actually I did not get the biggest one but 2 others similar ones are present). At 99.9% I think they should be (I just did not check). When searching for channels, I use a table in which the multiplier value is calculated according to the amount of data. Thus I take a channel as an existing one at 5% exceeding a confidence interval of 99% (I multiply this interval width by 1.05)

But as it turned out from the picture presented by Valadislav, it selects on purpose only those samples which begin at extremums of the channels besides a direct search of all channels in a row. Hence, I think there is some discrepancy between what we get and what Vlaidislav gets. I am not sure yet that the selection of channels specifically by swings may give some additional forecast accuracy, but the search algorithm will become more complicated - this is for sure. So far I haven't got around to checking it. Although, logically, it certainly makes sense - all traders do so - they draw these swings on their terminals and look at them all the time. But of course, there may be some ambiguity about the largest channel. It is only Vladislav himself who can tell if the largest channel really fits, or if it is the best one that could be taken at that time for the existing swing?
 
<br / translate="no"> But as it turned out, judging by the picture presented by Vlaidislav, in addition to the frontal search of all channels, he also selects those samples that start at channel extrema on purpose. Hence, I think there is some discrepancy between what we get and what Vlaidislav gets. I am not sure yet that the selection of channels specifically by swings may give some additional forecast accuracy, but the search algorithm will become more complicated - this is for sure. So far I haven't got around to checking it. Although, logically, it certainly makes sense - all traders do so - they draw these swings on their terminals and look at them all the time. But of course, there may be some ambiguity about the largest channel. Only Vladislav can tell us if the largest channel fits in his terminal or it is the best one that could be taken at the moment for the current swing.


Assuming that the market is fractal, the channels are fractals (this is what Vladislav is getting at). That is, larger channels contain smaller ones, and so on. He does not look further than 3 nesting levels (I am writing this from memory). Roughly, we build on daily charts, and in this channel, we build finer ones (4-hour or hour charts, and so on (though daily and 4-hour charts are not suitable due to time invariance))
Construction from an extremum... The main thing here is the idea, the calculation methodology will follow. It may be built from any stove, the main thing is to understand when to build. There should be a criterion for breaking one channel and starting to build another.
 
<br / translate="no"> Rosh 13.06.06 22:57
.....
ZS At what price did you draw the channels ?


I have tried it with (H+L+O+C)/4, but have also tried it with Close, I have not noticed any difference and have settled on this variant.


solandr 13.06.06 22:58
.....
I also using 99% probability did not get exactly the same channels as Vladislava (or rather did not get the largest - and the other 2 are similar present). At 99.9% I think

Same story. The largest (or rather very similar to it) was obtained only through commenting out the following line
if(CKO2/3>CKO)

SZZ Solandr, do you mind explaining the word "swing" in your last post means "peak" or something else (I've never encountered this term in FOREX and mathematics, unfortunately).
 
SZS Solandr, could you explain the word "swing" in your last post means "peak" or something else (unfortunately I haven't met this term in relation to FOREX and mathematics).

I have not met any clear definition of the word either, but I've got the following understanding from the context. As far as I understand, the swing in Forex is understood as a line connecting maximal and minimal at some interval. That is, I suppose that each line of ZigZag indicator is a swing. Rosh already gave a picture of ZigZag indicator in this thread. If I am wrong, correct me.
 
Dear Yurixx,

As I said in the quote above, "Whoever started this thread has absolutely nothing to say."<br/ translate="no"> Which is exactly what you have confirmed, having safely fled and made room for others.


The reason for my temporary withdrawal, I have already mentioned, is the writing and defence of my Diploma.
I've published my strategy test results a little earlier.

As for my opinion, I have it.


Rather, it's not you who has an opinion, it's you who has an opinion.
A little time has passed and in case you have forgotten, I want to remind you of your words:
And that it is a delusion, I assure you with all responsibility!


Dear Yurixx, I would even say dear!
Let me give you a piece of advice, before you assure anyone of anything, especially
.
with all responsibility!


You need to be 100% sure yourself, otherwise you
risk making a fool of yourself.

I thank you for OPENING this thread.
You started this thread, didn't you ? Come on, don't deny it,
we all know it was you. Thank you again for that!
It's no small effort to register a thread.
It's all thanks to Vladislav who not only made the system,
but also felt free to share his ideas with us.
And also thanks to everyone who took part in the discussion.


Dear Yurixx, I am amazed at your ability to twist and lick everyone nicely
and most importantly you are pretty good at it.


Sincerely,
Alexei.
 
Alex Niroba, stop it, it's not nice and it doesn't make you look good in any way. Trying to humiliate and insult another is not elevating yourself.
 
Dear Alex Niroba, I don't want to insult you in any way, but as a third party I would say that I take Yurich's posts much more seriously than yours. And I would be very grateful to you if you do not clog up the thread with remarks against other people, you'd better say something on the merits of the issue.
1...414243444546474849505152535455...309
New comment
Reason: