Elite indicators :) - page 324

William Snyder
9496
William Snyder  

!!xxxxxxxxxxxxxxx!!

mktsagli
35
mktsagli  

thank you MLADEN,

exactly what i was asking for

best regards

Manu
81
Manu  

mr tools,

I didn't ask for any code, just an answer : possible or not. But I understand your point of view.

mladen, sorry for being that impatient.

regards,

Mladen Rakic
151619
Mladen Rakic  

Tradefx1

No need for apologize

Your question reminded me of one thing that we (people) usually neglect.

Let me explain : in your question there is a part that is saying "when the price comes back near to a moving average ...". Now, as you know, a computer has to quantify everything, and that part would be translated to something like this :
if (previous absolute distance > 1 pip and current absolute distance <= 1 pip do something)
And there we come to the core of the problem :
why 1 pip and not 1.1 or 1.0001 or anything like that (or what to do with JPY crosses, or Mexican peso, ...)

why just previous value and not a lot of values

what to do when it is obvious that the price is "dancing" around that line (obvious to human, of course) - how to quantify that

and so on ...

And all sorts of similar questions. Once I have read a comment that we (human) can make a robot that can build another robot, but we can not make a robot that can tie show laces. The neglecting part in humans is that we do neglect the quantity of our implied knowledge and logic : what seems to be simple and perfectly logical when told in words, is, in some cases, impossible (still) to translate into computer code in an acceptable manner. And as you may have noticed, I did not answer that part of question - I simply think that there is no acceptable solution for that kind of problem. And I do hate to write long explanations

regards

Mladen

Tradefx1:
mr tools,

I didn't ask for any code, just an answer : possible or not. But I understand your point of view.

mladen, sorry for being that impatient.

regards,
Mladen Rakic
151619
Mladen Rakic  

An idea :

Majority of indicators use n previous bars to calculate current value of something. Like ATR : it takes last n bars and calculates what is the current average true range. But what if we took ATR from 24 hours ago center it (so get an ATR that was half periods before and after) and use that as a kind of "prediction" (so, "it was like this around this time yesterday, it will probably be similar today" kind of "prediction") See the comparison of the "predicted" (upper) and the regular one
It seems to give some results that can be used (especially when we take into account that ATR is normally used in EAs) , but let see if it really is so. As I said, just an idea

Also added an option to center it or not. If it is not centered then it calculates an ATR that was after the bar 24 hours ago (so it calculates what happened after that bar in that case, not before it as a regular ATR does)

PS: it is an intra-day indicator (it will not work on periods > 1 hour)

Manu
81
Manu  

mladen,

Thank you for your answer, I understand what you said and I see that my questions are going a bit nonsense in the coding environment. But I had to ask because as you know I am not a coder (I don't wanna tell you what's my job, cause you'll laugh) . I try to do my best to at least code my EA's inspiring myself of your work. I know it's obvious for you and surely ennoying to answer those questions, but it's my only way to learn about it.

Now you enlighted me about those limits, and I thank you by the way cause it's a very good thing, but keep in mind I will still bother you with dumb questions.

Mladen Rakic
151619
Mladen Rakic  
"but keep in mind I will still bother you with dumb questions"

How does a virtual (internet) murder look like ________________________

Seriously now, in my opinion, main purpose for posting source code is not to let others have the source itself (scammers will find their way with or without the source) but to let everybody see the code told in coding language (since it is still a language) and to eventually learn a new language and that way learn to communicate with that dumb thing called computer.

As of your job : some very serious studies are showing that best coders are the one that never had formal coding education. The reason is simple : they do not have constraints that the coders are usually having, and then, unconstrained, they usually ask the most important questions. So, just keep on asking

regards

Mladen

Tradefx1:
mladen,

Thank you for your answer, I understand what you said and I see that my questions are going a bit nonsense in the coding environment. But I had to ask because as you know I am not a coder (I don't wanna tell you what's my job, cause you'll laugh) . I try to do my best to at least code my EA's inspiring myself of your work. I know it's obvious for you and surely ennoying to answer those questions, but it's my only way to learn about it.

Now you enlighted me about those limits, and I thank you by the way cause it's a very good thing, but keep in mind I will still bother you with dumb questions.
Jimmy
652
Jimmy  

Trix

mladen,

Would it be possible to add a histogram (the difference between fast and slow trix) to the attached Trix indicator so it looks like the image.

thanks,

jim

Files:
Mladen Rakic
151619
Mladen Rakic  

Jim

Here you go
regards

Mladen

94315jim:
mladen,

Would it be possible to add a histogram (the difference between fast and slow trix) to the attached Trix indicator so it looks like the image.

thanks,

jim
Files:
Jimmy
652
Jimmy  

thanx mladen.........