Welcome to the Traders Laboratory Forums.
Forex Trading Laboratory Discussion forum for Forex traders - all forex pairs

Reply
Old 09-25-2007, 12:50 AM   #177

PYenner's Avatar

Join Date: May 2007
Location: New Zealand
Posts: 511
Ignore this user

Thanks: 0
Thanked 8 Times in 8 Posts



Re: Playing with the VMAR`s open research

Quote:
Originally Posted by walterw »
What I dont understand is what data he uses when you load the indicator from scratch again vs the real time thing... why does it get so slopy on rt ? dont want to sound annoying... hope I dont spell any bad word here... jejeje

Bruce, would you be interested to see what I am doing on paralel with cci ? and my ideas about a possible ultra modern oscillator... just ideas here, dont want to be annoying... anoying people get others anoyed and then the anoyed want to get the anoyer anoyed too.. so this anoyed people are so anoyed they just anoyedly keep anoying each other... isnt all this phrase anoying jejeje ... I think today trading got me anoyed ¡¡¡ cheers Walter.
On historical data it gets only one price to read per bar, the close price.
On a live chart, a 5 pip candle has 5 choices per bar.
So on a live chart it is following the ticks, faster than the bars,

The slowest line will often be 2 or 3 bars later, to the right.
One reason for attempting a binary version was to minimize the time delays that are built into the vma formula.
The binary version has no vma to slow it down 2-3 bars, but so far it doesnt do long HEs either...

You dont annoy me Walter, MT4 does, lots of things frustrate.
I need to get some trading done, make some $, coz nothing but indicators for a month. When I have a live trade and do other stuff, I can forget about the trade!!!! So need to do a bit of both.
The first problem I have with CCI is that it seems to be unsymmetrical or non-linear or something, when the price goes up in two steps the same size CCI shows the first rise big and the second rise looks small, danger danger non-linear not good.
PYenner is offline  
Reply With Quote
Old 09-25-2007, 01:42 AM   #178

PYenner's Avatar

Join Date: May 2007
Location: New Zealand
Posts: 511
Ignore this user

Thanks: 0
Thanked 8 Times in 8 Posts



Re: Playing with the VMAR`s open research

Quote:
Originally Posted by shreem »
Hello PYenner, after trying more harder and harder, still do not understand how to construct a VMA

I really want to be able to have this FantailVMA on Ensign Software but I need to understand how can I add a variable moving average in my chart before being able to add it to the ADX.

If you dont mind, can you share with me how you come to construct the VMA. What is the formula to create it?

Thank in advance and sorry if it appear as a dumb question as I am really not well versed in computer programming

Sincerely

Shreem
We all learn programming languages as we need them.

All of the MT4 files are txt files, you can open them with Notepad or word processor or Meta Editor in MT4 and try to follow what the code is saying.

VMA is very similar to the formula for an ema.

Sma formula can look like this-
SmaResult[]=(1-k)Oldprice[] + k*NewPrice[]
[] means a storage array, one price or result is stored for each bar on the chart.
k is called the weighting, in this case it is fractional between 0-1.
When k=1 you get SmaResult[]=NewPrice[]
When k=0 you get SmaResult[]=Oldprice[]
When k=0.5 you get an average,EmaResult[]=0.5*Oldprice[]+0.5*NewPrice[].
So its an average, that one looks like a simple average of prices from two bars, 50:50.

The Sma formula can also look like this, but now k is no longer fractional here-
SmaResult[]=((1-1/k)Oldprice[] + NewPrice[])/k
(This k behaves like a bar count).
k=4 is like a 4 bar average because
SmaResult[]=3/4 * Oldprice[] + 1/4 * NewPrice[]
The result is 3 parts OldPrice and 1 part NewPrice similar to a 4 bar average.
This k is often in integers, but there can be decimals like k=2.5.

Ema can give more weight to recent data- (with k fractional)
EmaResult[new]=(1-k)EmaResult[old] + k*NewPrice[new]
This is a running or moving average, the numbers stored in EmaResult[] can be mostly older prices (slow average) or mostly new prices (fast average) depending on the weighting k. k=1 gives fast average, stays close to new prices, this is like laddering. K is fractional here.
k=0 gives no change with new prices, like horizonal effect.

So the Chande VMA formula looks like Ema formula-(with k=VI fractional)
VMA[new]=(1-VI)VMA[old] + VI*NewPrice[new]
VI is the input signal, it provides variable weighting.
VI=1 is laddering, ignores old prices, follows new prices.
VI=0 is horizonal, ignores new prices and stays with-
VMA[new]=VMA[old]

Bemac used a 2 bar ema of the Chande formula, VMA4 uses the Chande formula with three different emas applied to it. Yeah it got more complicated.

But for feeding a signal into the Chande VMA formula, the input VI= "Volatility Index" has to move between zero and one. So many forms of indicator are not suited to giving 0-1 output and something has to be changed to give 0-1 output for sending to the VMA.

Last edited by PYenner; 09-25-2007 at 01:57 AM.
PYenner is offline  
Reply With Quote
Old 09-25-2007, 03:17 AM   #179

PYenner's Avatar

Join Date: May 2007
Location: New Zealand
Posts: 511
Ignore this user

Thanks: 0
Thanked 8 Times in 8 Posts



Re: Playing with the VMAR`s open research

Walter

Use FantailVMA4Turbo and FantailVMA3 as indicators to compare #3 with #4.
Dont use FantailVMA4 as an indicator, it is not indended to be added as an indicator, it is only there to make fantail lines.

I started this confusion sorry.
Better luck with 4Turbo.
Attached Thumbnails
Playing with the VMAR`s  open research-turbo4versus3.jpg  
PYenner is offline  
Reply With Quote
Old 09-25-2007, 06:05 AM   #180

Blu-Ray's Avatar

Join Date: Nov 2006
Location: England
Posts: 508
Ignore this user

Thanks: 164
Thanked 288 Times in 105 Posts



Re: Playing with the VMAR`s open research

Quote:
Originally Posted by walterw »
would you be interested to see what I am doing on paralel with cci ? cheers Walter.
Walter

I would be interested in your thoughts on incorporating cci into this method, as I, like yourself, am a big fan of cci.


Cheers

Blu-Ray
__________________

“ Search is the ultimate expression of the power of the individual, using a computer, looking at the world, and finding exactly what they want ” – Eric Schmidt, Google
Blu-Ray is offline  
Reply With Quote
Old 09-25-2007, 07:28 AM   #181

PYenner's Avatar

Join Date: May 2007
Location: New Zealand
Posts: 511
Ignore this user

Thanks: 0
Thanked 8 Times in 8 Posts



Re: Playing with the VMAR`s open research

How to use the three VMA4 files-

The files are used in a different way from the past in order to have the 3 vma lines with or without the fantail.

To use the fantail-
First add FantailVMA4.tpl as a template,
then add FantailVMA4Turbo.mq4 as an indicator to get the vma lines.

To use just the vma lines without a fantail-
Add FantailVMA4Turbo.mq4 as an indicator.

Don't add FantailVMA4.mq4 as in indicator in either case as it will only give you the first yellow line of the fantail...

Last edited by PYenner; 09-25-2007 at 08:09 AM.
PYenner is offline  
Reply With Quote
Old 09-25-2007, 08:07 AM   #182

PYenner's Avatar

Join Date: May 2007
Location: New Zealand
Posts: 511
Ignore this user

Thanks: 0
Thanked 8 Times in 8 Posts



Re: Playing with the VMAR`s open research

Quote:
Originally Posted by Blu-Ray »
Walter
I would be interested in your thoughts on incorporating cci into this method, as I, like yourself, am a big fan of cci.
Blu-Ray
Blu-Ray
I will attempt to "translate" VMA4 into VT type statements that may be more useful for others to follow than the MT4 version.
But I won't be attempting to actually code it on VT.

I ended up doing three vma lines so it got more complicated instead of less complicated.

I still have more to do on vma, months maybe.
Although vma4 seems like the first "trader friendly" version, I feel like I have only got to the start of understanding how it works, certainly not near the end, not nearly mastered yet.

Would also like to get an excel version running, but where is all the time supposed to come from?
Cheers
PYenner is offline  
Reply With Quote
Old 09-25-2007, 08:17 AM   #183

shreem's Avatar

Join Date: Jul 2007
Location: Montreal, Canada
Posts: 99
Ignore this user

Thanks: 228
Thanked 18 Times in 13 Posts



Re: Playing with the VMAR`s open research

Hello PYenner, thank you very much for your excellent explanation about how to construct the VMA. This is very helpful and will I start from there in trying to code it for Ensign Software.

You are right, coding certainely take a lot of time and it is a big learning process all by itself.

Sincerely

Shreem
shreem is offline  
Reply With Quote
Old 09-25-2007, 08:23 AM   #184

walterw's Avatar

Join Date: Nov 2006
Location: Argentina
Posts: 2,228
Ignore this user

Thanks: 0
Thanked 199 Times in 127 Posts



Re: Playing with the VMAR`s open research

Great explanations Bruce ¡¡... I know we all suffer here from the same problem "lack of time"... forget about cci... I think vma optimizations as you are doing is the way at this point.


Blu-Ray: some time today I will post some charts with the cci adding... cheers Walter.
__________________
you must enjoy trading... otherwise you shouldnt trade...
walterw is offline  
Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Last Post
Playing the Initial Balance TinGull Market Profile 13 10-19-2011 10:48 AM
Walter`s First Forex Research (various ideas) walterw Forex Trading Laboratory 157 09-13-2011 05:14 AM
Playing the Opening Gap Soultrader Technical Analysis 19 05-23-2007 05:10 PM
Playing Inside Days Soultrader E-mini Futures Trading Laboratory 12 03-06-2007 05:10 PM
Research On Forums and Communities. Comments Please. RichardKen General Discussion 6 02-18-2007 08:48 PM

All times are GMT -4. The time now is 03:33 AM.
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.
CS to VB integration by DeskLancer
©2006-2011 Traders Laboratory, All Rights Reserved.