EMA weight calculation - page 2

New comment
 
eddy >>:
"Часто вместо доли используют период, из которого рассчитывают эту долю: k=2.0/(1+period)." а 2.0 это тогда что, если не доля? ведь по твоим словам вместо доли используется период

Sorry, I can't communicate on these topics with people whose level of mental development is below the 3rd grade of secondary school. Have a nice day.

 
Mathemat, at least you answer. did you understand the pig? I didn't, and I pointed out exactly what
 
I cannot understand the logic of the function
 
k=2.0/(1+period) is your share.
If you chose period 9 (which is actually the period of a roughly equivalent regular waving), the fraction will be 0.2. The larger the period, the lower the share of Close[0] in the calculation.
If you don't want to deal with periods, use formula directly, specifying fraction at once. It seems to me that specifying a period for EMA is somewhat unnatural.
 
EMA[i] = pr*Close[i] + EMA[i+1]*remainder pr

pr=25.0/(1+period)

if the period is 4, and the cloze is say 4, 3, 2 and 1, then
EMA = 4*25/5 + ema[pred]*(1-25/5)
 
Mathemat >>:
Если ты выбрал период 9, то доля будет равна 0.2.
I can see this in the formula, but I don't understand the function and its meaning
 
Mathemat >>:
k=2.0/(1+period) - это и есть твоя доля.
Если ты выбрал период 9 (на самом деле это период примерно эквивалентной обычной машки), то доля будет равна 0.2. Чем больше период, тем ниже доля Close[0] в расчете.
Не хочешь разбираться с периодами - используй формулу напрямую, задавая сразу долю. Мне вообще кажется, что задание периода для ЕМА - это несколько неестественно.

Naturally, what is unnatural! ))) This is how it is done in MT and even more distorted - the period for EMA is set only as int type.

Meanwhile, the MACD, for example, has quite fractional periods if precisely recalculated from the author's coefficients. So does DiNapoli's MACD. The differences are significant.

My understanding is that this was done for MA uniformity. But in Metastock, for example, no one prevents the use of non integer periods.

 
Mathemat >>:
если период 9, то доля будет равна 0.2. Чем больше период, тем ниже доля Close[0] в расчете.

yema=fraction from cloz0 + "1-fraction" from the yema of the previous one

how is it a 1-share?

 
eddy >>:
это я вижу по формуле но не понимаю суть и смысл производимых в функции действий

In a formula, I should think. Not in the f-i. Where did the conversion formula come from is unclear? Or the exponential nature of the EMA?

 
It's not clear to me what the yema is all about. how it's calculated... it's added to the closing share of the cloz0... what?
12345678
New comment