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.
Reading this thread makes me what to rip my shirt off and hit something.
And really, I think most members have no idea what these guys are talking about anyways. The subject is way advanced for the normal guy. There is probably 5 people in this whole forum that can understand the subject. Most just want to hang around so they can download some magic cycle predictor indicator and make a gazillion dollars.
__________________
Life is a Gamble. You win some. You lose some. And life goes on.
I believe (guessing) the rest of the members wants to learn about Digital Filters than research about some diploma true or false or read last comments about the private life of somebody or whatever that involves only a minor percent percent of current thread readers .
IMHO.
Yes,you are right,consequently, I have edited or deleted most of the irrelevant info.
NOW...if you analyze it,the issue is VERY intriguing:
1-The guy preached about M1 predicting...BUT...He predicts on H4+H1..He should explain why this discrepance,there are many newbies that can be misleaded by his assertions.
2-Whenever he has been asked anything of relevance,like showing us a prediction based on M1,which will be extremely interesting...He switches gears to a totally unrelated issue,like my Diploma ...This is not the first time he "curtain smokes" the thread when he is asked some relevant questions...First it was accusing everybody and his uncle of having commercial interests,Last time it was "Arrows" and Comedy time,now it is Diplomas...I wonder what will come next...anything but a specific reply.
3-For months he has been here asking questions about other people`s methods..To make him post anything about his own we have had to practically extract it from him with forceps....Actually,if you realize it,he has posted more about MY methods(breaching his confidentiality agreement with a group of hard working people,and showing his true value as a person)Than about HIS...
Reading this thread makes me what to rip my shirt off and hit something.
And really, I think most members have no idea what these guys are talking about anyways. The subject is way advanced for the normal guy. There is probably 5 people in this whole forum that can understand the subject. Most just want to hang around so they can download some magic cycle predictor indicator and make a gazillion dollars.
Try this one,its name is Cycle_GJH4,it should be used in GBPJPY H4 timeframe,but you can test it on GBPUSD H4 and probably on H1 both pairs it can help you too...This is the red one at both pics.
Since most pairs are extremely correlated I won`t be surprised if it worked for a few of them too,but that was not its purpose...
1-It is based on MESA and is a composite of the 3 more stable cycles found at GBPJPY H4.
2-Doesn`t repaint
3-It is very robust,it works well but it is not perfect since it hasn`t been overfitted,see how it went out of sync for last 2 weeks of February,so...
4-..The way to trade it,as shown in one of the pics,is to ALWAYS confirm with Trendlines,then when both Cycle_GJH4 and the TL confirm a turn,you take this a your trend direction filter and trade in that direction with your usual methods ...This is just an additional tool in your toolbox.
5-This is like a knife,it can help or it can hurt you...use with care,play with it in demo until you are confident about it...then use it or discard it under your own responsability.
6-Always be aware of the higher timeframe,what can kill a trader using cycles is the trend,which is nothing else than a Higher timeframe cycle.
Just be carefull not to get repainted spectrum or maybe you want to
make SSA/HP casulal ??
Krzysztof,
you seem to be obsessed by the repaint issue. I guess you must have lost a lot of money with a misplaced interpretation of what a non-causal filter was telling to you.
Any time you analyze data, as you have a new tick arrival, the whole analysis changes. And it can be a big change. If you are doing spectral analysis, whatever the method you are using, be it mesa, fft, goertzel or whatever, every new tick changes the analysis itself. I mean, the whole world is repainting every millisecond, do you know greek philosopher Heraclitus? He says "panta rei", everything changes.
And that's what market is: change.
So the real problem for a trader is how to make money with change of price as time flows. And I believe that the first important thing is to put the right frame to the temporal horizon, as what is probable in the next 2 seconds is different from what is probable in the next 2 days. That's why SIMBA is right at pointing out your 'timeframe mess'.
Then,
if SSA or Hodrick-Prescott non-causal filters, which are not telling you to buy or sell, although your NN software may understand so , help in getting a sharper peak analysis with mesa or goertzel, like they do (see pics : on the left 'raw prices', on the right 'denoising with ssa'), I'm not scared of using them.
Krzysztof,
you seem to be obsessed by the repaint issue. I guess you must have lost a lot of money with a misplaced interpretation of what a non-causal filter was telling to you.
Any time you analyze data, as you have a new tick arrival, the whole analysis changes. And it can be a big change. If you are doing spectral analysis, whatever the method you are using, be it mesa, fft, goertzel or whatever, every new tick changes the analysis itself. I mean, the whole world is repainting every millisecond, do you know greek philosopher Heraclitus? He says "panta rei", everything changes.
And that's what market is: change.
So the real problem for a trader is how to make money with change of price as time flows. And I believe that the first important thing is to put the right frame to the temporal horizon, as what is probable in the next 2 seconds is different from what is probable in the next 2 days. That's why SIMBA is right at pointing out your 'timeframe mess'.
Then,
if SSA or Hodrick-Prescott non-causal filters, which are not telling you to buy or sell, although your NN software may understand so , help in getting a sharper peak analysis with mesa or goertzel, like they do (see pics : on the left 'raw prices', on the right 'denoising with ssa'), I'm not scared of using them.
Rich,
Congratulations for your very detailed analysis
By looking at your graphs we can see there are 3 or 4 near constant cycles,each one of them seemingly having small variations around the individual "nominal period"/"nominal frequency" axis...Is that so?
Krzysztof,
you seem to be obsessed by the repaint issue. I guess you must have lost a lot of money with a misplaced interpretation of what a non-causal filter was telling to you.
Any time you analyze data, as you have a new tick arrival, the whole analysis changes. And it can be a big change. If you are doing spectral analysis, whatever the method you are using, be it mesa, fft, goertzel or whatever, every new tick changes the analysis itself. I mean, the whole world is repainting every millisecond, do you know greek philosopher Heraclitus? He says "panta rei", everything changes.
And that's what market is: change.
So the real problem for a trader is how to make money with change of price as time flows. And I believe that the first important thing is to put the right frame to the temporal horizon, as what is probable in the next 2 seconds is different from what is probable in the next 2 days. That's why SIMBA is right at pointing out your 'timeframe mess'.
Then,
if SSA or Hodrick-Prescott non-causal filters, which are not telling you to buy or sell, although your NN software may understand so , help in getting a sharper peak analysis with mesa or goertzel, like they do (see pics : on the left 'raw prices', on the right 'denoising with ssa'), I'm not scared of using them.
Beautifull pictures. No I'm not obsessed with repaint, if you know what you doing that is fine you must be just carefull as I wrote. Also posting such pictures without explanation is very misleading.
Regarding mess. There is no mess at all here. I gave prediction using H1/H4 to align with Simba methods and also posted the link which proves that stabile cycles on M1 exist. Two independent things.
Any time you analyze data, as you have a new tick arrival, the whole analysis changes. And it can be a big change. If you are doing spectral analysis, whatever the method you are using, be it mesa, fft, goertzel or whatever, every new tick changes the analysis itself. I mean, the whole world is repainting every millisecond, do you know greek philosopher Heraclitus? He says "panta rei", everything changes.
I believe you should be more precised. Every new tick will change the analysis
related to FUTURE and CURRENT data but in case of SSA/HP will change HISTORY data in your system. How your system will act on this that's another story. Or I'm wrong ??
By looking at your graphs we can see there are 3 or 4 near constant cycles,each one of them seemingly having small variations around the individual "nominal period"/"nominal frequency" axis...Is that so?
Regards
Simba
You are right. The first row of peaks from the left is starting at 54 and slowly shifting towards 65 as you can also see in picture without the 3d elevation. The second peak is really very stable near 39-40 for all the observations. Keep in mind that I am using a relatively short window of 300 bars and that from the first observation (lower row of 2d pic) to the last (upper row) there is a 200 bar shift, that means that the peak remains stable for a time lapse of the same order of the window of obervation of past bars.
I also attach a text file with all peaks/valleys from the start to the end of simulation.
I believe you should be more precised. Every new tick will change the analysis
related to FUTURE and CURRENT data but in case of SSA/HP will change HISTORY data in your system. How your system will act on this that's another story. Or I'm wrong ??
Krzysztof
I think you missed the point.
Obviously past data cannot be changed by new data. But if you analyse the past with a new datum in your eyes, you can see another story, i.e. your analysis can be very different, whatever the analysis method.
So, a single new price tick can change the perspective of your analysis, we see it every day with price respect to S/R. A single new tick can break a support and change your perspective.
In the end, if you preprocess raw data with a high cutoff frequency non causal filter (in the pictures SSA smothes raw data with the equivalent of a 9 period lowpass), you are changing the whole input data of your analyzer, as you say, but, respect to what you are looking for (peaks in the range between 18-20 and 100), it is a small variation while the advantage it can give you can be way bigger (sharper peaks).
The demonstration is that analysis made with or without preprocessing the data leads to identical quantitative information in the band we are looking for (where the peaks are ), with a qualitative improvement (peaks in the interesting bands are sharper, peaks for small and untradable periods are smoothed)
...This is not the first time he "curtain smokes" the thread when he is asked some relevant questions...First it was accusing everybody and his uncle of having commercial interests,Last time it was "Arrows" and Comedy time,now it is Diplomas...I wonder what will come next...anything but a specific reply.
or maybe you want to 
Linear Mode
