What makes an unsteady graph unsteady or why oil is oil? - page 11

New comment
 
Urain >>:

Вы наблюдали за построением бара в тестере?,

открытие если клос меньше опен те бар в целом медвежий то тики идут сначало к максимуму потом к минимуму а потом к клосу(имеется в виду тики М1),

в реале это невозможно никто не знает как закроеться бар.

А вот советник случайно выявил эту закономерность постоения тиков в тестере.

It still does not make sense. First, even in the tester, the EA does not know how the bar will close.

And secondly, if the tester reverses the direction of the ticks, the grail won't work?

 
Andrei01 wrote >>

It's still not logical. First of all, the EA does not know how the bar will close in the tester.

And secondly, if the tester reverses the direction of ticks, the grail will not work?

The EA does not know, but the tester does, and the EA simply uses the tester's knowledge.

If the speed (tick size) is higher than the threshold, the system catching the breakthrough will be activated and if the bar is going to be big (and the tester knows about it in advance), then first the strong push against the movement is directed and then the system attracts and successfully catches profit on the opposite movement.

 
а... I can see if it's pipsqueaky now. :)
 
NTH писал(а) >>

Well, all TA is certainly not built on that. Predictions are probability, and with some logic it can be eliminated and a pattern detected.

Take triangles, pennants, fibos, or any indicator-based stuff. The occurrence of something like that allows people who believe in it to predict the direction of prices. There is another category that looks for such a combination and tries to substantiate statistical significance of such a figure with the help of a tester. The meaning does not change.

 
timbo писал(а) >>

God be with you, where did they go? In non-stationarity, simply put, at least one of the parameters is not a constant. And scientifically speaking, it's https ://ru.wikipedia.org/wiki/%D0%A1%D1%82%D0%B0%D1%86%D0%B8%D0%BE%D0%BD%D0%B0%D1%80%D0%BD%D0%BE%D1%81%D1%82%D1%8C.

I don't want to get into it. I agree, there are, but variables, although if defining stationarity one should not stop at moments, but rather give my BP model and within it the definition of stationarity or non-stationarity, like Box and Denkins for example.
 
lea писал(а) >>

An example, please.

Signal ~ fair value?

What does fair value have to do with it, there is no fairness at all. I wrote about U-turns.
 
faa1947 писал(а) >>
I wrote about reversals.

And I asked for an example of stationarity in markets.

 
The price minus the muving is quite mean reverting. I won't speak for stationarity.
 
Mathemat писал(а) >>
The price minus the mouving is quite mean reverting. I won't speak for stationarity.
I wrote above that wavelet decomposition is very useful, when the first column of the matrix is the trend, the second column is the average of the difference between the trend and BP, etc.
 
Prival >>:


а меня гнет другой вопрос и не по детски. поясню. Анализ тиков и сразу пипсовка это вилами по воде писано. Второе меня просто не устраивают бары. бары это сжатие информации с потерями и потери необратимы. кто то берет 1 цифру клоуз 5 минутки. за час у него 12 цифр. посчитайте на досуге через частоту дискретизации, какой минимальный приод колебаний возможно обнаружить.
уменя в анализе тот же час, только оцифрован он правильно. Данных намного больше... выводы делайте сами.

Well, it's worse than just a loss of information. It's a distortion that causes something to appear in the spectrum that wasn't there. And the inability to separate the original components from the artificially created ones. One can only estimate the error.

And the unchildish question that interests me-who says it can be predicted at all?

позволяет точно спрогназировать допустим движение величиной 160-170 пунктов (5 знаков)

Does your prediction work that way?

Or is the key word here "suppose"?

1...456789101112131415161718...34
New comment