Welcome to Forex-TSD!, one of the largest Forex forums worldwide, where you will be able to find the most complete and reliable Forex information imaginable.
From the list below, select the forum that you want to visit and register to post, as many times you want. It’s absolutely free. Click here for registering on Forex-TSD.
Exclusive Forum
The Exclusive Forum is the only paid section. Once you subscribe, you will get free access to real cutting-edge Trading Systems (automated and not), Indicators, Signals, Articles, etc., that will help and guide you, in ways that you could only imagine, with your Forex trading.
Elite Section
Get access to private discussions, specialized support, indicators and trading systems reported every week.
Advanced Elite Section
For professional traders, trading system developers and any other member who may need to use and/or convert, the most cutting-edge exclusive indicators and trading systems for MT4 and MT5.
What I did only is take the EURUSD 4h data and generate the spectrum using DFM.
I get the following peaks: 12, 21, 30, 39 and 72
First, I don't know which peaks are significant or not? How do we know it?by their height,the 2 significant ones are 30 and 73 in your pic
Then using those informations, I should be able to build a digital indicator. If for example I want to create the SATL indicator, how should I proceed? How do we set the parameters P1, D1, A1, Db and so on? I guess we have to use the cycles' values to filter one cycle but I don't understand what each parameter is doing..P1=73..D1=30..A1=40..dB=0.08..BASICALLY YOU ARE TAKING ALL CYCLES BETWEEN 30 AND 73 AND SMOOTHING THEM TO CREATE A NOISE FREE(lower noise actually) MODEL OF PRICE..leave A1,Db at their predefined settings you only have to worry about P1 and D1Also, if I want to create a PCCI filter, should I set the parameters so that the indicator is not crossing the mid-line too often and when it does then it is showing a valid move or reversal?P1=73..d1=30..Try it,leave A1 and Db as they are preset
How do we create and use the filters' sets? What is their interest?1-TO FILTER NOISE IN REAL TIME WITHOUT DELAY(with minimal delay)..so,for example ,a change in the slope of satl signals a change of trend,while a change in slope of a 73 periods sma has a delay equivalent to half its period approximately,36periods delay.2-To make oscillators based on half the period of the dominant cycle,in your pic,the dominant cycles are 30 and 73,so,if you use RSI,Stochastics,etc..you will probably get very good results by setting their period to 15...and 36
I know that that makes a lot of questions and I'm looking forward to your answer.
Regards,
Daniel
Hi Daniel,my reply is above,why don`t you just do the SATL and PCCI and post them here so we can take a look and see if they need some modifications?I explained to you one of the ways to do what you requested,usually it is the most useful one,but there are alternatives in case the end result doesn`t models the price as it should.
Hi Daniel,my reply is above,why don`t you just do the SATL and PCCI and post them here so we can take a look and see if they need some modifications?I explained to you one of the ways to do what you requested,usually it is the most useful one,but there are alternatives in case the end result doesn`t models the price as it should.
Regards
Simba
Hello Simba,
Thank you very much for your reply!! I'll create a SATL and a PCCI indicator and post them so that you can check if they are looking good.
If I want to create the SATL and FATL indicators, I think I've to use the short periods for FATL and the long periods for SATL. Am I right? Is it ok for example to take P1=73 and D1=30 for SATL and P1=12 and D1=21 for SATL?
And in this case, I should not divide the values by 2, like for a normal stochastic or RSI indicator?
About the filters' sets, how do we combine the filters if I want to create a STLM - FTLM indicator for example?
Thank you very much for your reply!! I'll create a SATL and a PCCI indicator and post them so that you can check if they are looking good.OK,let`s see what we can combine If I want to create the SATL and FATL indicators, I think I've to use the short periods for FATL and the long periods for SATL. Am I right? Is it ok for example to take P1=73 and D1=30 for SATL and P1=12 and D1=21 for SATL?Yes,it is ok,this is the right concept.
And in this case, I should not divide the values by 2, like for a normal stochastic or RSI indicator?No,absolutely not.You only divide values by 2 when you use them in traditional oscillators,basically,for example,if cycle is 50,half cycle is 25,if you use a williams % range oscillator,,for example,you want to fit it to 25 periods,the half cycle,so it will give you the highest percentual readings at the top of the cycle(the top of the UP half),and the lowest values at the bottom of the cycle(the bottom of the down half).When you use noise filtering devices(SATL,FATL,even a traditional sma) you want to filter OUT the noise outside the dominant cycles,so you use the full cycle(s) period(s),never the half cycles...do a visual check..use a 73 periods sma delayed by 36 periods..and a 30 periods sma delayed by 14 periods..and check if they represent the past price history accurately or not..probably they will,only issue will be that you will have a lag,since smas will end 36 and 14 points ago....that is why digital filters are very useful..their lag is minimal.
About the filters' sets, how do we combine the filters if I want to create a STLM - FTLM indicator for example?This is more advanced,stlm2&ftlm use double buffering to achieve additional smoothness,let`s stick to the basics first,once you know how to do a customized satl or a fatl,we can proceed,and then it will just be a matter of copying and pasting,onceyou understand the concept and the filter mechanics,it becomes easy.
Regards,
Daniel
Hi Daniel,as usual ,the reply is above ..Regarding STLM2 and FTLM,I will explain once you can manage SATL and FATL..anyway,I have to tell you that STLM2 and FTLM are already very robust indicators (even more than SATL),and it is very difficult to find a pair and timeframe where it is worth the effort to build a customized STLM2/FTLM,anyway,we will do it if you want..as a next step,once you can handle SATL and FATL as second nature.
Regards
Simba
So, let’s try to build one of those digital filters.
First I export the 4h data of EURUSD and perform the spectrum analysis in DFM.
According to the picture, I get the following peaks at about: 12, 19, 29, 41, 63 and 88.
Now, I’ll try to generate some SATL and FATL indicators. For the SATL, which should be slower and on a longer period, I will filter the cycles of the peaks 41, 63 and 88, so I’ll use 88 P1 and 41 for D1 and leave A1 and passband ripple at they default values.
After looking at the chart, I choose to filter using P1=63 and D1=40 as it gives me a curve with sharper turns easier to see.
For FATL, I’m using P1=31 and D1=12 to make a nice filtered curve for the short cycles.
For the PCCI indicator, I’ll try to keep longer cycles, as for SATL. After some tries and trying to get a cross of the zero line when there is a change in the trend, I choose P1=88 and D1=44 for the PCCI filter.
That's it! :-) All the filters are in attachment for control :-)
So, let’s try to build one of those digital filters.
First I export the 4h data of EURUSD and perform the spectrum analysis in DFM.
According to the picture, I get the following peaks at about: 12, 19, 29, 41, 63 and 88...OK
Now, I’ll try to generate some SATL and FATL indicators. For the SATL, which should be slower and on a longer period, I will filter the cycles of the peaks 41, 63 and 88, so I’ll use 88 P1 and 41 for D1 and leave A1 and passband ripple at they default values.
After looking at the chart, I choose to filter using P1=63 and D1=40 as it gives me a curve with sharper turns easier to see.OK,better this way
For FATL, I’m using P1=31 and D1=12 to make a nice filtered curve for the short cycles.Why those settings?19 and 12,29 and 12or even 29 and 12 could be testable..but why 31?
For the PCCI indicator, I’ll try to keep longer cycles, as for SATL. After some tries and trying to get a cross of the zero line when there is a change in the trend, I choose P1=88 and D1=44 for the PCCI filter...OK ,YOU CAN TRY AN ALTERNATIVE ONE FOR 63 AND 29/31..and compare
That's it! :-) All the filters are in attachment for control :-)
Hi,now the most interesting part...please read my comments above and be so kind to post some charts where you put the different digital filters on the pair and timeframe that you selected for analysis..Nothing like seeing the results on an actual chart
Hi,now the most interesting part...please read my comments above and be so kind to post some charts where you put the different digital filters on the pair and timeframe that you selected for analysis..Nothing like seeing the results on an actual chart
Hi Simba,
Here is the chart with SATL, FATL and PCCI.
I've tried with the values you gave me and it was ok for FATL and PCCI. I had no reason for taking 31 instead of 29
I've also noticed that depending on which value I'm using for D1, the curve can be offset compared to the price. Is it a bug in the DFM tool?
I'm longing to know more about the other more complex indicators we can create.
Here is the chart with SATL, FATL and PCCI.OK..looks like you have a way NOW to define the trend for EURUSD H4..for example PCCI above 0 plusFATL above SATL and SATL slope is positive..looks like a good h4 uptrend that you can trade with oscillators,RBCI,STLM2,etc in h1,m30 or m15
I've tried with the values you gave me and it was ok for FATL and PCCI. I had no reason for taking 31 instead of 29 OK
I've also noticed that depending on which value I'm using for D1, the curve can be offset compared to the price. Is it a bug in the DFM tool?NO..it is just the fact that you are changing the cutoff period..try to do an alternative SATL with p1=63..d1=29 and compare with your actual one,it will probably be faster but less smooth...but probably useful too..try to do it and post a pic and compare
I'm longing to know more about the other more complex indicators we can create.Yes,understandably so...and you will realize that most of them are based on SATL and FATL
Hi Daniel,
As usual,my comments are located above...try to decide which of the SATLs and FATLs suits better your trading style,and then we will proceed to STLM2 and FTLM..but first,we will need to stop at RSTL and RFTL.To understand the concepts behind RFTL & RSTL it is important that you read my explanation about centered moving averages at the SimbaConMan thread,or,even better ,alternatively,read the J.M Hurst book.
for some reason i cannot get the data that i export out of metatrader (csv) to load in the spectral analyzer...i get an error....can someone walk me through that process?
thanks,
cl
edit- i figured it out nevermind
Last edited by clahn04; 01-27-2008 at 02:19 AM.
Reason: i figured it out :-)
As usual,my comments are located above...try to decide which of the SATLs and FATLs suits better your trading style,and then we will proceed to STLM2 and FTLM..but first,we will need to stop at RSTL and RFTL.To understand the concepts behind RFTL & RSTL it is important that you read my explanation about centered moving averages at the SimbaConMan thread,or,even better ,alternatively,read the J.M Hurst book.
Hi Simba,
I've read the whole SimbaConMan thread. I've seen the centered moving averages method. It looks really interesting. What is the relation between that method and RSTL and RFTL?
I've read the whole SimbaConMan thread. I've seen the centered moving averages method. It looks really interesting. What is the relation between that method and RSTL and RFTL?
If you read the Hurst book,you will realize that the difference between a cycle and itself delayed by half the cycle period,in a perfect theoretical world,will nail the tops and the bottoms of the cycle..in real life,it does a fairly good job to approach tops and bottoms,the problem being the time delay inherent in the use of centered(delayed by half) moving averages,and the fact that cycles are non stationary,unless you can find some cycle with very high Bartels.
To solve the first problem problem,we can use a SATL and a SATL delayed by half its period(RSTL)and estimate the difference(that is what the STLM2 does,BTW),so we have a "conceptual Hurst " method in real time...and now the question..check the standard SATL and the standard RSTL..Is it delayed by half the period?
If I want to create the SATL and FATL indicators, I think I've to use the short periods for FATL and the long periods for SATL. Am I right? Is it ok for example to take P1=73 and D1=30 for SATL and P1=12 and D1=21 for SATL?Yes,it is ok,this is the right concept.
Linear Mode
