Spearman_Rank_Correlation, 4 timeframes matrix
The first time I used, it seemed the closest to the 'holy grail' for me. Need the tms.mq4 included in the rar file for the matrix to work. Bear in mind that it is not a repaint indicator but a mtf one, so the higher timeframes recalculate till the time period closes.
Comments
Very funny matrix indicator
so, the entries and exit are located when 4 vertical red or blue dots are shown
Exactly jag... that simple!
I am thinking that by using some kind of extrapolation for the higher frames maybe we have something interesting here.
Did you know, that the Pearson correlation is also very interesting ?
correlation between SP500 and EURUSD by John Last
http://beathespread.com/blog/view/2727/actual-correlation-between-sampp-500-and-eurusd
A study for selecting the best indicators
http://beathespread.com/file/view/18662/study-correlation-indicators-to-price
http://beathespread.com/file/view/2737/pearson-correlation-rank-indicator
I made also for beathespread a study about thie Pearson correlation and lade a selection of the best indicator able to fit the curve
http://beathespread.com/file/view/15749/pearson-correlation-for-several-pairs-gt-for-stevegee58
The simple SPearman- Rank correlation is here
http://beathespread.com/file/view/2737/pearson-correlation-rank-indicator
I just noticed that the "TMS basic.mq4" is nothing else than the "SPearmanRankcorr.mq4"...
Let us firstly remember that financial time series are NON-STATIONARY for the reasons we have discussed , and NON-LINEAR in that they are non-Gaussian.
So, in this case, the Spearman correlation can be used for fetching entries and exit. Try just to attch the SpearmanRankCorr.mq4, and you will see instantaneously it
Correlation is an idea that arises from LINEAR algebra. To ask if two non-stationary time series are correlated is like asking "Who married the fridge?". The question is quite meaningless. Asking if two non-stationary series are CO-INTEGRATED is quite a different matter and a highly valid question.
If you ask this, you can REALLY learn something about the future!!
But how to mesure the Cointegration ? if someone could help us , it will be great
You guys are on to something here. Keep at it :)
It repaints on the next three higher timeframes......so how we can enter?
You can't, until it is stabilized. Too "repainty" for now.
How I can see when it will stabilize?
Virux84,
it gets stabilized when Jag and Kerdos has found the way to stabilization ;)
I think it is just a fancy tool unless you define simple rules and test them in an EA. Then you will have some idea if it is worth or not.
I think most of the forums are creating bad habits (making them loosing money) by convincing people that what they see is real. Most of the time it is not.
So you see a good looking idea, the right approach is to see, yeat it is coll, let formalize the rules, code them and test them.
This is one way of doing things another way is to focus on global macroeconomic issues and see if your statistical tools confirm the analysis.
It is complicated because people mix: Digital signal proceccings, then they mix it up with statistics, then comes the practical implementations necessitating coding skills. And last but not least data mining.
Those are completely different fields of knowledge and quite a few are good in everything. So there are more amateurs than you can think even among professionals, who do not even know that they are amateurs.
Or it is just lazy Fridey afternoon monologue LOL.
You can check also this discussion.
http://beathespread.com/blog/view/2802/correlation-or-cointegration#comment_2052