Searching for market patterns - page 44
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Are you normal? There's an awful lot of doubt.
From the podium of another planet. WHAT'S NOT CLEAR?
Fuck you - live funny.
You can't drink that much. There's still a star to be seen.
Why is that?
And the code is really not cleaned up, I would say horrible, difficult to understand. In addition, there are errors in order modification.
But even though it is "semi-working", it worked on the section I gave it.
the filter response to a single impulse is predefined for some time, but after the price impulse it is not certain that the price behavior will follow the same rules as in your picture (I meant the rules. of course it will not follow the rules as in the picture)
unless you make a lot of such pictures with each pair and different periods and watch the result - maybe you will see common cycles or something like that.... the only thing I see useful is the same scaling for all pairs for further calculations - it's like a shift to relative prices.
This is exactly what I was talking about. The information is lying in front of your nose and no one can see it.
The filter's reaction to the real prices may look like anything, it's not what's important.
The area where the filter does not react to the input signal is highlighted in red. In this picture it is 400 bars, which is a lot. In this case we had 0 at the input and applied 1. We could feed -1, 1000 or -1000, this section will look the same and the differences will show further. In this section there is 0 at the output only because there was 0 at the input before. The 400 bars at the filter output are fully determined by the past data. Now let it not be 0 at the input, but actual prices to which the filter reacted in some way. This section will generally look somehow different, it will probably be a curve, but still at 400 bars it will be fully determined by the past data. By feeding any crap I can get an accurate filter response for 400 bars ahead, and it will always be that way no matter how the price goes next. Neither polynomial nor any other extrapolation gives that kind of range and prediction accuracy, not even close.
Now someone will probably say that the filter extrapolates as much as it lags, so such tricks will not yield anything. Try it and you will see the following interesting things. Or don't, like here...
This is exactly what I was talking about. Information lies in front of your nose and no one can see it.
The response of the filter to real prices can look anything, that is not what is important.
The area where the filter reacts almost nothing to the input signal is highlighted in red. In this picture it is 400 bars, quite a lot. In this case we had 0 at the input and applied 1. We could feed -1, 1000 or -1000, this section will look the same and the difference will be shown further. In this section there is 0 at the output only because there was 0 at the input before. The 400 bars at the filter output are fully determined by the past data. Now let it not be 0 at the input, but actual prices to which the filter reacted in some way. This section will generally look somehow different, it will probably be a curve, but still at 400 bars it will be fully determined by the past data. By feeding any crap I can get an accurate filter response for 400 bars ahead, and it will always be that way no matter how the price goes next. Neither polynomial nor any other extrapolation gives that kind of range and prediction accuracy, not even close.
Now someone will probably say that the filter extrapolates as much as it lags, so such tricks will not yield anything. Try it and you will see the following interesting things. Or don't, like here...
my avatar is counting down...
when i'm sad, i remember this article http://intervist.narod.ru/statia1.html. we'll meet again in a billion years anyway)
you got a mosquito's eye there!
I don't need students. (Highly) Not yet born.
I liked the title "search for market patterns"...(and as if answering myself) yes, the patterns are common:
1.Entry points (= opening a position) in the selected area of the chart should be a lot. After all, if there are few...well, for example, 3 (three), it may be a fluke, and this is roulette.
2. Stop-loss and Take-profit should be related to each other by some coefficient.
3.The trading system should include "uncertainty", that is, stop-loss and take-profit should not "start to control the position" immediately at the moment of opening, but after some time.
And then just look for where your system deviates from the entry point more often... that's it, for fuck's sake.
you've got a mosquito's eye in there!
what what eye???
I'm reminded of an idea... It's based on the dollar index principle, i.e. each pair linked to it is a certain percentage.
So it's exactly the opposite - subtract the impact of the dollar from the movement of the pair. The idea is to see only purely the movement of the currencies themselves .....
So just thinking out loud on Fridays ))))
with a beer)))