Fisher - page 11
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The lying Yurk version
Hi, Devil2000 you are right, the Fisher zTransformation is a very valuable analysis method, but the Yurk-Version of the fisher indicator is a lying version without practical value.
The Yurk-version contains a programming error: when calculating its averages it is stepping back from presence to past, thus pesenting as past results what nobody could know at that time.
I correted this mistake, creating the version Fisher_m10 of this indicator in the thread https://www.mql5.com/en/forum/trading_systems.
I attach a screenshot where both versions (downloadable on my homepage http://home.arcor.de/cam06/fisher/) are compared. There you can see how the Fisher_Yurk version shifts its values back into the past.
The Yurk version for example indicates today "we have down trend" and tomorrow (after refreshing the chart or close/open Metatrader) it shows the present time as "up trend". So this inddicator version is useless.
Kind Regards
Martin
-------
Hi, Devil2000 you are right, the Fisher zTransformation is a very valuable analysis method, but the Yurk-Version of the fisher indicator is a lying version without practical value.
The Yurk-version contains a programming error: when calculating its averages it is stepping back from presence to past, thus pesenting as past results what nobody could know at that time.
I correted this mistake, creating the version Fisher_m10 of this indicator in the thread https://www.mql5.com/en/forum/trading_systems.
I attach a screenshot where both versions (downloadable on my homepage http://home.arcor.de/cam06/fisher/) are compared. There you can see how the Fisher_Yurk version shifts its values back into the past.
The Yurk version for example indicates today "we have down trend" and tomorrow (after refreshing the chart or close/open Metatrader) it shows the present time as "up trend". So this inddicator version is useless.
Kind Regards
Martin
-------So you're saying from your graph that Yur4ik's version shows a downtrend?
On your attachment Yur4ik's chart hasn't signaled any trend change.
Both charts remain up. It doesn't prove anything.
So you're saying from your graph that Yur4ik's version shows a downtrend?
On your attachment Yur4ik's chart hasn't signaled any trend change.
Both charts remain up. It doesn't prove anything.His chart proves where the uptrend 'should' begin and where it moved back to.
The easiest way to prove that it repaints is to mark the changes with vertical lines on the 15 or 30 minute chart and then come back a day later and you will see how it doesn't line up.
the lying indicator version
So you're saying from your graph that Yur4ik's version shows a downtrend?
No, that's not what Im saying.
What I pointed out is the fact that the Yur-Version of the fisher indicator contains a heavy programming error.
If you don't believe, read the code.
Yurk's loop 'for(int i=0; i<Bars; i++)' is stepping back from 0=presence to 1,2,3,... into the past, calculating i=past results with present data.
That's simulating future knowledge.
As I saw this fact has already been pointed out in this thread several times for example by Emerald King
(https://www.mql5.com/en/forum/173169/page2) .
The result of this programming error is the 'repainting behaviour' of this Yurk version already mentioned in this thread too.
The Yurk-Version cant detect turming points earlier than the Fisher indicator without programming error, but afterwards Yurk repaints the past as it would have known what would happen. See attached screenshot.
Besides the screenshot in my post I gave an example what consequences this programmig error can result in:
It can happen that this indicator makes you buy today and when you've lost all your money and try to recall the situation next day this indicator has repainted itself and tells you the lie it would have advised to sell yesterday and not to buy.
Kind Regards
Martin
I think the problem is there are about 5 different versions of this around.
Fisher_Yur4ik_v2 repaints terribly. If it's painting over any signal then yeah...it's crap. And the alert version (wherever that came from) doesn't work at all.
The version I have does a pretty good job though.
Gramski.
Feb2006 is right. Every "repaint" indicator are useless in real time trading. But you mention before that your version doen't repaint it's color, why don't you post it here so we can observe it?
Feb2006 is absolutely correct.Manytimes i watched the Yur4ik_2 indicator without blinking my eyes.It is a nice lying indicator.
Try this one...
Tell me what you think...
working version
A first look at the 'Fisher_Yur4ik_Test.mq4' code shows that it has got the same wrong stepping back mechanism as version 'Fisher_Yur4ik_2.mq4'.
But you have to differentiate between the effects of Yurk's programming error and the value of the Fisher zTransformation itself.
The Fisher zTransformation is a quite powerful method and as far as present results are concerned the results of the Yurk's version are right in a certain sence.
Yurk's stepping in the wrong direktion causes past bars to be wrong only.
In regard to the present bar his wrong direction has got another effect: his average calculation is reduced to one single bar, you are presented the raw value.
If you know what you do, using this raw value can be absolutely an advantage, it may be very erratic but it is very fast too.
But to know what you do, means you must be able to check with a realistic picture of the past and just this purpose is hindered by the redrawing behaviour of Yurk's version.
So I suggest use the attached version without programming error, set PriceSmoothing=0 and IndexSmoothing=0 and you get a raw value too, but you can see (in past) what you do.
Kind regards
Martin
A first look at the 'Fisher_Yur4ik_Test.mq4' code shows that it has got the same wrong stepping back mechanism as version 'Fisher_Yur4ik_2.mq4'.
But you have to differentiate between the effects of Yurk's programming error and the value of the Fisher zTransformation itself.
The Fisher zTransformation is a quite powerful method and as far as present results are concerned the results of the Yurk's version are right in a certain sence.
Yurk's stepping in the wrong direktion causes past bars to be wrong only.
In regard to the present bar his wrong direction has got another effect: his average calculation is reduced to one single bar, you are presented the raw value.
If you know what you do, using this raw value can be absolutely an advantage, it may be very erratic but it is very fast too.
But to know what you do, means you must be able to check with a realistic picture of the past and just this purpose is hindered by the redrawing behaviour of Yurk's version.
So I suggest use the attached version without programming error, set PriceSmoothing=0 and IndexSmoothing=0 and you get a raw value too, but you can see (in past) what you do.
Kind regards
MartinThat's right, the historical plots are hindered. You can only test yur4's out in real-time.
Thanks for that (indicator).
Gramski.