On the unequal probability of a price move up or down - page 131

 
khorosh:

On the real quotes, the picture is not impressive. Here is the publicly available NMA :


Show it not on 40 counts, but on a few hundred. Otherwise the scale of comparison is different.
 
Mikhael1983:
Show it not on 40 samples, but on several hundreds. Otherwise the scale of comparison is different.

something similar to yours can be found at https://www.mql5.com/ru/blogs/post/732070

HMA is a rather artificial filter with no theoretical basis, but a similar result (and with less computation) can be obtained in other reasonable ways.

There is an arithmetic nuance - if a filter (average, derivative, whatever you want to call it) has a period/depth/scale, then its graph will revolve around the corresponding SMA. And this SMA will be a good approximation, because it will be the limit.

О прекрасном
О прекрасном
  • www.mql5.com
Ранее представленное и несколько раз анонсированное "прекрасное" :-) Как и обещал выкладываю индикатор. Между прочим третья попытка за пару часов - то ли сайт глючит, то-ли firefox. Я блин код быстрее свёл в одно целое Итак, индикатор раскладывает цену на три фазы, как в электрике : параметр только один - PERIOD . Если вы считаете что время от...
 
Mikhael1983:
Show it not on 40 counts, but on several hundred. Otherwise the scale of comparison is different.

2 days.

 
Alexsandr San:

It's nice!

- Give me that indicator, I'll evaluate it for you in real time.

https://www.mql5.com/ru/code/10272

HMA
HMA
  • www.mql5.com
Просмотров: 18988 Рейтинг: Опубликован: 2011.04.27 08:18 Обновлен: 2014.04.21 14:55
 
khorosh:

2 days.

There you see. Nothing outstanding anymore, just the same as the price, with a slight smoothing. Well, mine can in the same way at 2 days with a reduction in the number of summands to come close to the price:


 
Mikhael1983:

Let's increase the number of addends in the filter sum from the last picture:

It's prettier already, isn't it?

Increase it some more:

khorosh, can your Jurik do that?

it's just that the height of the candles has decreased

 
Renat Akhtyamov:

it's just that the height of the candles has decreased

don't stop the man from going to Taylor's row and the component methods
 
Mikhael1983:

Let's increase the number of addends in the filter sum from the last picture:

It's already more beautiful, isn't it?

Increase it further:

khorosh, maybe your Jurik is like that?

Jurik no. But AMA, if you pick up the parameters, could be similar. An adaptive algorithm that changes the averaging period depending on volatility.


 
khorosh:

Jurik no. But the AMA, if you pick the parameters, could be similar. An adaptive algorithm that changes the averaging period depending on volatility.

This is a dead end. There are no averaging periods in my filter (when calculating the filter value in reference i, the information is at worst 1 reference in the past, and that's rare). Otherwise it would be impossible to do such a thing:

 
Mikhael1983:

It's a dead end. There are no averaging periods in my filter. Otherwise it would be impossible to do that:

Well, that's a bit of a mouthful. I remember now that you have written about your filter before.