Walter`s Forex "Trend Trades"

[PYenner]

Hi Walter

A picture of a first attempt at a turbo version.
Not sure if it will be too extreme.

Black is Fantail1
Yellow is turbo
Ignore Red, its just the switching signal to the VMA so I can see whats happening.

Try not to keep you waiting much longer but need some sleep.
Cheers

[walterw]

	Quote:	PYenner
	Hi Walter A picture of a first attempt at a turbo version. Not sure if it will be too extreme. Black is Fantail1 Yellow is turbo Ignore Red, its just the switching signal to the VMA so I can see whats happening. Try not to keep you waiting much longer but need some sleep. Cheers

Pyenner, you rock man ¡¡ that relationship looks very clean and leading at the same time... I believe we will get to the point of getting to an "almost" perfect timing... great stuff ¡¡ cheers Walter.

[PYenner]

	Quote:	walterw
	Pyenner, you rock man ¡¡ that relationship looks very clean and leading at the same time... I believe we will get to the point of getting to an "almost" perfect timing... great stuff ¡¡ cheers Walter.

Yes sharper but also the HE's tend to be shorter, it may need different lines in the fantail for timing. May be confusing at first. But 1-4 bars faster so worth a try...

[walterw]

	Quote:	PYenner
	Yes sharper but also the HE's tend to be shorter, it may need different lines in the fantail for timing. May be confusing at first. But 1-4 bars faster so worth a try...

well about HE`s ... if this particular version would be used for timing only we should be ok... the idea actually on this optimization is to get clean and leading entries... so let the HE from the previous version keep working for us in terms of settuping... I believe we will get to the point of using diferent versions for diferent definitions... cheers Walter.

[walterw]

Here I attach the template of the 1 min chart for my last possible timing optimization... (blue and white lines)

Now I believe that what Pyenner is programming will mostly outperform this one... so stay tunned... cheers Walter.

[PYenner]

	Quote:	Clym
	I have attached my attempt to duplicate the rainbow in MS excel in hopes that someone who has a complete understanding of the formulas may be able to see where I off and help me get it working correctly. I would like to get a full understanding of the calculations. I worked from Walters post (Bemac's code for VT platform) Clym

Clym
That is amazing.
Had no idea Excel was up to that sort of effort.
The black line of the fantail should be hugging the Close price.
It looks like the final formula has to be wrong for so much distance to accumulate between the black VMA line and the Close price.

This is the MT4 code for the Chande VMA.
VarMA[i]=((2-Const)*VarMA[i+1]+Const*Close[i])/2;
In MT4 [i] is the current bar and [i+1] the previous older bar.

The ADXR option can be deleted.
The 100*Wilder (3 locations) serves no purpose, delete the 100* or leave it, makes no difference.

EmaIndex and VarMAPr are counterproductive, they tend to put a slope on the flat area, delete them.

VT Wilder function uses a 15 bar ema when 8 bars are called.

Diff=HHV(_ADX,8)-LLV(_ADX,8) and
Const=(_ADX-LLV(_ADX,8))/Diff
are all important, make sure the conditional rubbish is not interrupting data flow from _ADX to Chande. Where division by zero must be avoided, the default should be zero, no funny numbers.

This should probably be on a programming thread somewhere but I don't know where that might be. Good Luck.

[PYenner]

Clym
Maybe try the "Playing with the VMAR`s open research" thread if you need me.
PY

[PYenner]

Clym
Const should resemble the main output from an ADX.
The Plus and Minus things should mimic the ADX component signals.
If you can monitor those signals inside Excel it will make life easier.
I use a spare indicator line to display signals in chart, offset from the Close price, eg
Indicator[i]=Close[i] + signal[i]

[PYenner]

Clym
DL FantailVMA2.mq4 from the VMAR research thread and open it with Notepad, its a txt file. Its a faster version than Bemac but it may give some clues about signal flow.

[PYenner]

Clym
I don't understand Excel well enough.
Your Diff should not be stuck on 100.
But your ADX final has enough 100's in it to be causing the problem.
Your DM's and DI's look to have way too many zeroes in them.
It may be that your IF AND stuff is faulty.
Plus logic attempts to find rising trend and rising trend within increasing volatility.
Minus logic looks for falls and falls within rising volatility.
It is abnormal for both to be low at the same time.
In fact they should tend to stay above zero quite consistently.

TH and TL need to be displaying their decimals, need to see the pips.
TH TL and TR look okay but note how TR is rarely zero?
The DM DI stuff should look the same way, normally above zero.
---------
Maybe this will help [i+1] is older than [i] ok?

TR=TH-TL;//2 bar unnormalized price difference, comprehensive volatility measure of current bar, always positive.
//Includes any gap data between bars, and price jumps caused by missing bars.

//ADX looks for trends in Highs, and in Lows.
if(High[i]>High[i+1] && Low[i]>=Low[i+1]) PDM=High[i]-High[i+1];//2 bar unnormalized price difference, both High and Low rose.
else
if(High[i]>High[i+1] && Low[i]<Low[i+1] && High[i]-High[i+1]>Low[i+1]-Low[i]) PDM=High[i]-High[i+1];//Rise within a volatility increase.
else PDM=0;//Zero or positive result, zero case includes volatility decrease options.

if(Low[i]<Low[i+1] && High[i]<=High[i+1]) MDM=Low[i+1]-Low[i];//2 bar unnormalized price difference, both fell.
else
if(High[i]>High[i+1] && Low[i]<Low[i+1] && High[i]-High[i+1]<Low[i+1]-Low[i]) MDM=Low[i+1]-Low[i];//Fall within a volatility increase.
else MDM=0;//Note, zero or positive result,zero case includes volatility decrease options.