Playing with the VMAR`s open research

[walterw 10]

 

Midas is an amazing concept, getting rid of the time variable altogether and using volume in its place.

When I can get my head around it, I think that will answer some difficult questions for me that are all about reading the market. Could be very powerful if understood and exploited properly, you have a good eye for this stuff.

 

 

jejeje... yes Bruce, its certainly a Hidden Gem ¡¡ at least Paul Levine left that very nice educational material wich creates a very nice base to start from...

 

Now remeber to use hat man ... jejeejje... cheers Walter.

[homestudy 10]

Just saw them and wanted to try to help this great thread. I have been following it from day one. Hats off to you guys this work is amazing!

 

HS

[walterw 10]

Just saw them and wanted to try to help this great thread. I have been following it from day one. Hats off to you guys this work is amazing!

 

HS

 

 

Thanks a lot HS ¡¡ its nice to see this good will from you... cheers Walter.

[PYenner 10]

Walter

Question time.

I have only found eld vwap files and they are encripted so I cant read them.

Need an idea about what user settings are needed especially for starting or restarting at what(?) location.

 

eg Looks like jperl was using a 51 bar setting on 2min chart for volume distribution history so I assume vwap used the same setting, so if we use 1min chart, that means start on current bar and read previous 102 bars (2hrs) of data.

 

Alternative is to have a user setting for "start x bars back from current bar"

and sample "y bars of history".

 

What controls do you need to get the thing to start/restart on a pivot point or time or price or back/forward x bars, what???

[Blu-Ray 10]

PYenner

 

I'll send you a PM with the code of the ELD.

 

Hope this helps

 

Blu-Ray

[walterw 10]

 

Alternative is to have a user setting for "start x bars back from current bar"

and sample "y bars of history".

 

What controls do you need to get the thing to start/restart on a pivot point or time or price or back/forward x bars, what???

 

we could input date and time could we ?

 

PYenner

 

I'll send you a PM with the code of the ELD.

 

Hope this helps

 

Blu-Ray

 

I will apreciate that Blu...

 

 

I am smelling Midas here Bruce ? jejej cheers Walter.

[PYenner 10]

Blu-Ray

Thanks heaps.

 

Walter

The three platforms I run here use three different midnights, none is est.

Even the two that are close to GMT differ by an hour.

One seems to have different "midnights" for each pair, arghhh*#.

This is the United Nations here...

 

Date and time might be the best.

People will have to work out their own time offsets somehow if they want to use a common "midnight" or any other "time".

 

There is a hazard here with the way MT4 remembers your last settings, may have to find a way around it or... you go on holiday for a week, come back, start MT4 and it takes 2 hours to start because it is running 200 SDs per bar on a weeks worth of bars on every chart you left vmar running on, bit of a time bomb especially on 1min charts, we will see.

 

Yes Midas is calling me too.

Vwap looks like the "time diluted" version of Midas.

Excel may be needed for midas, got it but don't know how to use it yet.

Some Chimp is keeping me busy...

 

There is a possible problem with forex volume, the gap data, in between the bars. I do not know if the bar data includes the volume data from the gap. I only know the bar data does not include the price data from the gap ticks. I need to find the answer before I start using volume big time.

 

MT4 does not have crash protection against division by zero, it locks the computer, you have to reboot then your default template is corrupt and it doesn't even know what charts or indis you were running, uses the next best template it can find, big mess. Gunna have to be extra thoughtful with this stuff. Taking a simple approach could have nasty consequences.

[walterw 10]

Well dont go nuts Bruce... its clear that mt4 has its limitations... I dont want to distract you too much on this... you know we have more than enough good arsenal to trade fine at this point... you see if its worth it the time and work... thanks for your great disposition Bruce ¡¡ cheers Walter.

[PYenner 10]

Walter

I am strongly suspicious of statistics for predicting the future.

I have seen many times how support can become resistance but that does not mean that statistics had anything to do with it.

It will lose my confidence completely once it gets much past 2SD's.

Relying on 3 SDs to tell you anything reliable is unrealistic to anyone who understands SDs, but how many do?

 

It does look to have big plusses, that justifies it.

But I am fearful of encouraging misplaced trust in things that are not understood well enough by most people that they know when not to trust them.

[Blu-Ray 10]

Just to add something else into the mix , I've coded up the ADXVMA6 into a CCi.

 

Don't know if it will have any benefit, but you can play around with it, to see if you find it useful.

 

I've coded it up for TS and MT4 ( Don't hold your breath on the MT4 version :shocked:) as it's the first one I've got to work.

 

Test it and see..................

 

 

Cheers

 

Blu-Ray

ADXVMA6_CCI.ELD

ADXVMA6_CCi.mq4

[walterw 10]

Just to add something else into the mix , I've coded up the ADXVMA6 into a CCi.

 

Don't know if it will have any benefit, but you can play around with it, to see if you find it useful.

 

I've coded it up for TS and MT4 ( Don't hold your breath on the MT4 version :shocked:) as it's the first one I've got to work.

 

Test it and see..................

 

 

Cheers

 

Blu-Ray

 

Oh My Lord ¡¡¡ I had been expecting this indicator a good time ago ¡¡

 

I will test him Blu... tell you how he does, specially on ploting issues... thanks man ¡¡

 

what a team we got here ¡¡ now we also have Sparrow who started to program our indis for Ninjatrader platform ¡¡

 

The chimp is so happy that he is dancing ¡¡

 

 

 

cheers The Chimp.

[Blu-Ray 10]

Oh My Lord ¡¡¡ I had been expecting this indicator a good time ago ¡¡

 

I will test him Blu... tell you how he does, specially on ploting issues... thanks man ¡¡

 

what a team we got here ¡¡ now we also have Sparrow who started to program our indis for Ninjatrader platform ¡¡

 

The chimp is so happy that he is dancing ¡¡

 

cheers The Chimp.

 

 

hehe, another great video

[walterw 10]

Ok tested Blu-Ray`s ADXVMA6_CCI for mt4 and there is no plotting issues (repaintings)... works very good ¡¡ cheers Walter.

[walterw 10]

Hi Bruce... I need some help here...

 

Unicorn has coded his version of vma`s... actually I had been interacting with him on this thread here http://www.traderslaboratory.com/forums/24/walter-s-forex-vma-trend-scalps-2558-7.html arround some of his charts... they look very nice and I finded some interesting nuances to his particular charts...

 

he has been kind enough to share his code with us wich was done for amibroker charting platform...

 

could we code this for mt4 as well ? please dont go bald ¡¡ cheers Walter.

[PYenner 10]

Unicorn has coded his version of vma`s for amibroker charting platform... could we code this for mt4 as well ?

Will look into it.

[walterw 10]

Will look into it.

 

 

Thanks Bruce... I forgot to attach code... cheers Walter

vmar_1[1].afl.txt

[Armand 10]

what a team we got here ¡¡ now we also have Sparrow who started to program our indis for Ninjatrader platform ¡¡ The chimp is so happy that he is dancing ¡¡

cheers The Chimp.

 

 

Just an FYI...the Ninjatrader platform has a great dom (The same one I use with TransActFutures). It lets you place bracket orders, target and stops all with one click. You can also drag and drop your targets and stops. Great control. Thanks Sparrow....Armand

 

 

[PYenner 10]

Vidya3.mq4

This vidya was ported from the VT version by Amritend Maji (Maji).

The settings were for 5min chart.

Seemed to get the thumbs up for matching VT charts.

Chande invented this too, but there are variations around and uncertainty about which are true to Chande.

 

xpMA_v2.mq4

Universal xpMA by Codersguru.

Change MA_Type to get:

 

0 Simple moving average,

1 Exponential moving average,

2 Smoothed moving average,

3 Linear weighted moving average.

4 Double Exponential Moving Average.

5 Triple Exponential Moving Average.

6 T3 Moving Average.

7 Jurik Moving Average.

Vidya3.mq4

xpMA_v2.mq4

[Sparrow 10]

PYenner: the xpMA 4 and 5 settings are actually the other way around if you look at the implementation

 

Here's my first NT verison of unicorn's adaption of the FantailVMA1.

Hopefully someone can test it and tell me if it works fine.

 

Cheers

[PYenner 10]

PYenner: the xpMA 4 and 5 settings are actually the other way around if you look at the implementation

Sparrow

Thank you, it looks like you are right but I'm not going to mess with it for now. Users of xpMA be warned.

 

Sparrow, if you are also familiar with T3 this might be a simple question for you, if not just ignore it because I'm being lazy and looking for a quick fix.

I suspect that T3 was intended to use a sequence of ema's but a version I am looking at has used a variable where an array should have been used if ema's were the intended result. The difference is that arrays will work fine on both bars and ticks, but a variable will malfunction on live ticks.

This is what the code using variables looks like and cci is supposed to end up holding a T3 version of a cci input signal. Would e1, e2 etc make more sense...? W1 and W2 weightings are ema type weightings eg (n-1) and n.

 

for(int i = Bars - 1; i >= 0; i--)

{

cci = iCCI(NULL, 0, CCI_Period, PRICE_TYPICAL, i);

e1 = w1*cci + w2*e1;

e2 = w1*e1 + w2*e2;

e3 = w1*e2 + w2*e3;

e4 = w1*e3 + w2*e4;

e5 = w1*e4 + w2*e5;

e6 = w1*e5 + w2*e6;

cci = c1*e6 + c2*e5 + c3*e4 + c4*e3;

 

I am new to just about everything so I will lean on an experienced shoulder if I can. Don't bust a gut on it, just answer if its obvious to you.

Cheers

Lazy PYenner

[Sparrow 10]

PYenner, my knowledge of how indicators work internally is very limited and my adv. maths skills are lacking unfortunately.

I have a programming backgroung and just got into indicator development but I am a quick learner.

 

I might have an idea though because I am facing a similar problem at the moment.

My guess is that the value for the previous bar is stored in the variable.

If it's calculated on the close of the bar its value is correct but if you use every tick then it will represent the value of the last tick and not the ema value of the last bar.

 

Hope it makes sense.

 

Cheers

[PYenner 10]

Sparrow

Thank you.

You see it the same way I do.

Thanks for quick answer.

Bruce

[Sparrow 10]

Glad to help Bruce, unfortunately I'll have to patch some of my indicators although I personally seldomly use the intra bar feature.

 

Here's the NT version of the FantailVMA1 modified with unicorn's code again.

[Sparrow 10]

Believe me I am not doing this to drive up my postings count

Unicorn just modded the indicator, so I am trying to catch up.

Parameter for the stochastics period added to the NT version.

FantailVMA1_v1.2.zip

[unicorn 10]

Chande invented this too, but there are variations around and uncertainty about which are true to Chande.

 

Hello Bruce, Sparrow, Walter;

 

Chande has introduced the concept of varying the coefficient of the exponential moving average, f= 2/(n+1), by multiplying this coefficient by a volatility measure or by a trendiness measure, f_chande:

 

factor = f * f_chande;

 

vidya = Variable Index DYnamic Average, is the adaptive moving average (AMA) of the price, usually the Close.

 

vidya = AMA( price, factor);

which as you know is:

vidya = factor * price + (1-factor) * previous_value(vidya)

 

In his book, Chande uses among other measures the absolute value of the Chande Momentum oscillator (CMO), abs_cmo

 

cmo(n) = 2*rsi(n) -0.5; re-scaling of rsi in the range (-0.5, 0.5)

abs_cmo = abs(cmo);

 

factor = f * abs_cmo;

vidya = ama (C, factor);

So, this is the original formulation of Chande Vidya (to be precise, the original formulation that uses one of the 3 volatility / trendiness measures cited in the book).

 

now, abs(cmo) is very close to adx; as a matter of fact, Chande claims that it is superior to adx.

 

From this point, using stochastic_adx = stoch (adx (adx_periods), stoch_periods) is a very smart and natural evolution. :thumbs up::thumbs up::thumbs up:

I don't know who came up with this idea; I first saw it in Walter's reference of Bemac's fantail. THANK YOU Walter.

 

Hope this clarifies things.

 

Now, for the purposes of this vmar research, I had initially restrained stoch_periods to be equal to adx_periods, in order to reduce the number of combinations. After Walter's interest in the behaviour of the signal vma_1(3,3) with respect to the rainbow_base vma_1(4,8) I figured that he might need the option to adjust stoch_periods, so that he can look into different settings.

 

cheers

Unicorn.