Digital ACSTrend - page 7

 

MESA Mod

Here is an example of the mod of ASCTrend. I am still testing it because I am not sure as usual to call the right buffer. However I tell you exactly what I did. I used the Richcap library. He has prepared an indicator called cut R-MESA-cutoff_frequencies. And I use this indicator to adapt my jurik WPR (which in his turn is the heart of ASCTrend Signal indicator)

https://www.mql5.com/en/forum/173071/page35

If I am right and if it is so easy (however terribly heavy for the machine) we will have a new series of adaptive indicators based on mesa. We have now based on Cycle period indicator and Hilbert Transform. This would be again a hybrid technology between dynamic adaptation and jurik digital filtering.

However I am not sure I did it right on the screen shot. This is still a prototype.

Please note that when we use MESA adaptation we may expect razor signals when we have persistent time series. For antipersitent market conditions with a lot of pink noise do not expect the MESA to perform.

I use risk 8 to 10 with 8 of digital smoothing. You need to install everything from the link above. Take care in the default there is 0 backward bars so you need to set this manually otherwise you will have nothing on the screen. Please look at the thread Richcap made other useful things that may be used as a filter for this system as for example the adaptive SATL. We even may consider a full MESA mod. We are limited only by our imagination.

However keep in mind one principle. When we have a low phase singularity my theory is that all the systems tend to converge into one global solution. Whatever directional algorithm you will use you will get the same picture. The adaptive algorithms allow us to survive in very unclear territory, they do so by slowing and accelerating. Imagine a road, if the road is a motorway everyone is able to run fast. However if we have a mountain road we need to slow down. This is what the adaptive algorithms would do for us, and that is why they are useful. This is just a comparison.

Files:
 
John Last:
Remember the Simple Moving Average is hard to beat LOL.

This was the normal ASCTrend with its standard settings of Risk 3 and Money Risk of 1.5.

And on this shot is my version with Mama.

With the same settings but in addition we have

Fast limit 0.2

Slow Limit 0.05

Once again you see that when the FGDI is below the 1.5 line and is red (iVAR below 0.5) we have periods when we have nice signals (well we do not know if we will have a trend but during this time the directional algorithms perform at their best).

The best time to trade is when the FGFI makes a fractal break - out and is pointing down. When it is extremely low very often we have a Black noise reversal as it is predicted by the neural net that we can expect something like this.

what is the dotted indicator on your chart? Have your tried using your system based on renko/rangebars?

 

Dots indicator

Hi,

This is one of the most promising mods of ASCTrend

https://www.mql5.com/en/forum/180729/page2

I know it is time consuming but you need to know what is it. In fact it is the ASCTrend dots indicator but based on the MAMA indicator (the original is based on cross-over of SMA).

There are many different mods, because you can put whatever you want into ASCtrend (the indicator with the dots) with two lines of code.

We have with Jurik, with Alma, with Ama, with Mama etc. Maybe for the Elite section it can be done with All moving averages and so we can choose and the choice is not easy: 20 indicators. (As far as I know this is an elite indicator, and that would be the ultimate moving cross - over fans indicator).

But what we have here is enough for me.

Well I did not tried on Renko, I prefer the range bars. However I did not tried it yet. However I doubt for the use of the cycle period adaptation in timeless bars.

 

Thanks John for the answer !

Yes you are true, on renko chart it's not good, but on 1 hour charts Asctrend Mama do the job, that's sure !

:D

 

Fractal patterns

Here I add some fractal patterns on the 5 m chart. This is really useful for day trading and scalping. With this you will know how much potential has every signal.

When we have a low fractal dimension the movement has a potential. When we have a low fractal dimension the movement has not a profit potential (even consider not to take the signal given by directional algorithm).

Here on the shot I use two different colors manually to color the zone.

Have a look at the last signal, there it is a typical black noise reversal (when in low fractal dimension the reversals are abrupt and very sharp). This type of price action is typical for the low fractal dimension. We count on digital filters for that manually you can't take that trade (at least I can't LOL).

Here I add the JStochastic with MESA. Of course you need the JStochastic libraries here on the thread and the MESA library on the link in the previous post. This is another concept indicator.

 

Hurst exponent indicator faces a slow or no show?

Hello, John,

I enjoy your theory of fractual dimension breakout. I put 2 Hurst exponent indicators on eur-usd, daily chart to study their performance, and found:

1. FDGI(30) is bettern than iVAR(5).

2. Both failed to show the big trend of 2nd half of 2009 in early stage.

I believe A part and B part (see attached chart) have the same space fractual dimension in terms of eye view. My question is how can we identify and classify the fractual space and its dimension? (I haven't read any book about fractual and chaotic theory,yet).

Most of trend run out in a angle channel with variable angle and length. So, fractual dimenion tools may be needed to address this issue for a better indication of real fractual dimension change or fractual dimension transfer.

2nd half of 2009, eur-usd ran out of different type format of trend, But, it is big trend. This type of trend is often in the market place for all pairs. I don't know whether it is in low phase space singularity or in high phase space singularity? May there be a middle one: "mid phase spce singularity?

Enjoy your works. Best regards.

 

Where the trend starts?

Hi,

I will meditate more on your questions because I do not understand them in full.

However I will address some issues that I feel like understanding them.

1. First there is only an estimation of the fractal dimension. We use different algorithms and we try to estimate it.

2. You say that both indicators failed to see where the big trend began in 2009.

I will tell you that those indicators do not misguide you. And that is a part of the technical analysis fallacy. Please look at the beginning of the trend when the beginning of the down movement that was a part of the normal range for the up trend or at least can be explained by that.

3. When you go to a lower time frame you will get another picture of clearly persistent market movements.

4. The fractal dimension and its estimation is not connected with the trend.

Those two concepts can be confused but they are different. The trend is a concept of the technical analysis (higher highs or lower lows). The fractal dimension describes the fractal properties of the time series (how do they behave, do they look as a straight line or they are really jagged)

And we need to look at several time frames to get a picture.

This concept is very, very useful in the general Technical analysis, especially when you analyse a break - out of a trend line (a valid break - out requires a low fractal dimension).

5. I am not sure that the daily time frame is really the place where the fractal dimension can be used. I look at it at maximum 1 h time frame.

Another way is to use tic charts, because the time is not a constant not only in the the physics but also in the Forex. Some times when you will have a lot of volatility you will miss a lot of information in the current bar because the time gets accelerated.

6. I call it a singularity because the price action is really strange from the normal market conditions. When we have low fractal dimension at different key levels a lot of algorithms are able to find the same solution and different investors start to have the same horizon and the price is propulsed strongly in kind of a mono-waves movements.

And vice - versa for the high - phase spaces, it is really difficult the price start to make very complex ranging patterns. However it is difficult for me to have the support and resistance right.

7. Of course you can name a Medium Phase Space Singularity when there is a difference between the Fractal dimension and the Trend movement.

The possibilities are open. You are the author now of this concept.

Files:
daily_1.gif  34 kb
 

Multi time frame MESA adaptive Trend filters

Here I will add some indicators that may be usefull for trend filters. They use the indicators of Richcap MESA adaptive. I use for the multitimeframe as a carrier indicator one very famous indicator of Tudor Girl from the Forex Factory.

However they are really heavy for the machine and are not properly tested yet.

However I want to show that the Richcap MESA adaptive indicators should be considered very seriously.

You need to download the libraries that you can find on the following links:

https://www.mql5.com/en/forum/173071/page35

https://www.mql5.com/en/forum/173071/page35

 

Multitimeframe MESA Adaptive R-FTLM-STLM filters

Here I add the last of the collection of Multi-time frame MESA (Maximum Entropy Spectral Analyse) Adaptive momentums (R-FTLM-STLM).

This is a great achievement of Richcap. I wanted to make it more user - friendly using multi - time frame screening.

The indicators are there:

https://www.mql5.com/en/forum/173071/page35

You need also the MESA library, it is in the link in the previous post.

However this is very heavy. And it is default set to adapt every 100 bars. I do not think that makes any sense because I need a constant adaptation.

However doing so it crashes my Metatrader. I you make an adaptation of every 10 bars of the original indicator (not the multi frame) you will see a very, very good oscilator.

I added the parameter chooser. If you choose 0 you will have the Fast Trend Line Momentum. If you choose 2, you will have the Slow Trend Line momentum.

Do you see the vertical bars? The vertical bars are the periods when the adaptation is made. If you look at the objects you will see that there even are marked the used frequencies.

I do not know if it is possible to add more colors at the multi-frame indicator, for example 4 colors: above zero and increasing (slope), above zero and decreasing; below zero and decreasing; below zero and increasing

 

Colour mod

I was able to adapt another code. However I am not still able to make multi-time frame with 4 colors. This mod really need that.

However this mod is with bars not with line. I think that this is not only nice looking but useful. In fact I did an adaptation of every 10 bars. In the original settings the adaptation is default set to every 100 bars.

IMHO this is not enough. We have to adapt constantly.

So there is a parameter which will rectify the distortions, as they occur with the adaptation. With the bars as we have bars we are visually OK with the distortions no need for misleading rectification.

I used this intermediate code to insert into ready 4 colors multi - frame so that would save a lot of work. Need a little help.

Files:
Reason: