Breaking through the morning flat - which pairs? - page 20

 
baltik писал(а) >>

When optimised, all passes end in -0


EA does not support the situation when h1 is in the previous day.
Set h1 from 0 to 4, and you will be happy....
 
baltik писал(а) >>

Question (and all of the above) let me rephrase,
the indicator on which the Expert Advisor is currently running is not lying, it is not its fault?
Is it possible to increase profit factor using another indicator?
Just in visualization mode i-Regr.mq4 indicator looks more consistent and balanced.


It is good. But the screenshots show practically the same indicators. What is the trick with the second one?

 
Oper писал(а) >>
An acquaintance of mine had a problem with a midnight flop breakdown.
He just took it and shot himself.


Not for nothing. A flat, as far as I know, is when something is horizontal. So, it doesn't go up, so what? You can live with it.
Even when there is a long-term downtrend, it is not worth shooting. IMHO.
.................
In the future, I will try not to react to Oper's remarks, because this thread is alive.
 
ikatsko писал(а) >>

In the original Dserg-a indicator, the signals are generated immediately after a break in the channel. But the width of the channel had to be set manually. My suggestion is to set the expected time (range) for "catching" the morning flat. During this period of time, the indicator selects MINIMUM channel width going through various channel variants from StartTime to FinishTime and only at the finish (when there are no other variants) a channel with minimal width (r0) is drawn whose length is equal to the number of bars between Start and Finish (Nlin). And if by that moment the channel (obtained) has been broken, output signals are generated. I am now thinking about the criterion of channel length optimization. >> May somebody have some ideas?


Ivan, the ideas you put forward are certainly interesting. I am also thinking in this direction...

1) The following is not clear. The minimum width of the channel must be equal to the length (number of bars) = 2. So the minimum length is limited somehow?
2) What do you want to see as a result of optimization?
3) I set the t0 parameter = 1.618 and get these views. I even dread to look inside!
Maybe you have already fixed variants of this indicator?

 
lasso писал(а) >>


Good. But from the screenshots, the indictors are virtually indistinguishable. What's the second one's trick?


Visually: when 2 indicators are on the same scale, i.e. on one chart, we see different plotting
i-Regr.mq4 builds by candlestick body, not by shadow.
I cannot explain (it is in the code)... switching windows on different timeframes using i-Regr.mq4 does not slow down the terminal


The originals in better quality of graphics on p.18
Let's assume that at the moment we can visually see on the 7am screenshot that the channel is 3-4 pips narrower, the lower boundary of the channel ending is not suspended,
it is being built .
now let's analyze the situation when the EA has got into a losing situation
it was 21 times during the test since 01.01.10 and on all these days between 0-7 hours we have seen strong moves
if we suppose that the indicator, used in the Expert Advisor at that moment, builds the channel taking the noise (shadows) into account
then we get a channel not precisely (not really) synchronized with the market situation
The indicator I consider has no such an advantage (disadvantage) - it looks at the real price movement
Open and close and not cheated from high & low. That seems essential to me.

lasso wrote >>

EA does not support a situation where h1 is in the previous day.
Set h1 from 0 to 4, and you will be happy....


Thanks for the clarification, Vitaly. I wanted to experiment with the channel from 22 to 01.

 
baltik писал(а) >>


on the test since 1.01.10 is 21 times, and on all these days between 0-7 hours we see strong movement


This is already an argument. I'll buy a beer and give it a try.
But I don't think it's gonna work....

 
baltik >>:


Визульно: когда 2 индикатора в одинаковом маштабе а то есть на одном графике, то видим разные построения
i-Regr.mq4 строится по телу свечи, а не по теням.
Не могу объяснить (дело в коде).. переключение окна по разным таймфреймам с использованием i-Regr.mq4 не подтормаживает терминал


Оригиналы в лучшем качестве графики на стр.18
Допустим что в данный момент на скрине 7 утра визуально видно что канал на 3-4 пункта уже, нижняя граница окончания канала не подвешена,
а участвует в построении .
теперь рассмотрим ситуацию когда советник попадал в проигрышную ситуацию
на тесте с 1.01.10 - это 21 раз, и во все эти дни в промежутке с 0-7 часов мы видим сильное движение
если предположить что Используемый в данный момент в советнике индикатор строил канал учитывая шум (тени)
то и мы получаем канал не точно (не реально) синхронизированный с ситуацией на рынке
а индикатор рассматриваемый мной лишен такого достоинства (не достатка) он смотрит на реальное движение цены
опен и клоз и не подвергся обману от хай & лоу. Мне кажется это существенно.


Спасибо за разъяснения Виталий. Хотел поэксперементировать с каналом от 22 до 01.



The linear regression channel can be plotted by close or by high/low
The first option is implemented here.
 
Dserg писал(а) >>


The linear regression channel can be plotted by close or by high/low
The first option is implemented here.




You can add that to implement the second option, you need to type false in this place instead of true

      BoxH = iCustom(NULL,PERIOD_M15,"!LinRegrBuf",true,4*(h2-h1),2,0);
      BoxL = iCustom(NULL,PERIOD_M15,"!LinRegrBuf",true,4*(h2-h1),1,0);
 
baltik писал(а) >>

Indeed, changing UseClose = true|false results in this difference in channel width


 

At home was at hand only a DC with 5-digit quotes the result surprised
I did not expect such a catch in the tester to see 1 screenshot 4 digits - 2 screenshot 5 digits





Where did the profit go?!!! ;) Maybe that's where to start looking for a fix.