Please tell me the formula of the EMA indicator, or the indicator itself.

The bottom line is that instead of summing the unit value (with the corresponding coefficients), sum the values of the EMA of different periods.

For example, EMA1280, EMA320, EMA80, EMA20, EMA5

with the same increasing coefficients as the original. The highest is at the EMA 5, below the EMA 20, and so on. If the original meaning is that the last value of the bar is more important than the previous ones, because it reflects the latest prices, then here the last trend is more important than the previous trends. But, the previous ones with deletion increase in size and indicate a more general trend that will be taken into account in the formula.

## bar

All the same, the urdala_ma indicator is the best of the “cars”!

## nektomk

Ma2 Ma3 max is just a coefficient vector. with scales. You can add them (vectors) at once and not blow up the terminal’s brain with strange averaging and counting of individual ma-shek

well there is no fish there 🙂 and within will be..guess what ? 🙂

## qstr

urdala_ma

there are also ordinary cars. in the first post thought, take the original ma formulas with closets, add up on small different periods and see what happens.

ma2= (a1 + a2)/2

ma3= (a1+a2+a3)/3

ma2+ma3 = a1/2 + a2/2 + a1/3 + a2/3 + a3/3 = 5 * a1/6 + 5 * a2/6 + a3/3

can be divided into 2. ema coefficients are even more complicated, but the essence is the same: the near coefficients will increase more, the far ones less, and the last ones will add up without changes.

## bar

Sergey Rashevsky has a good indicator urdala_ma, it is much more effective than any EMA and works quite differently.

## avatara

## seriousracoon

Take the values of the machines, multiply each by the weight, sum and divide by the sum of the weights. What’s so complicated?

## arteinvolo

## capitalplus

You’re right. I’m sure this is another perversion with MA. I’m just curious to see her. I initially thought that when summing up the average value will be obtained, which on the chart will look like the same SMA, only with an averaged period (MA10+MA20) 2=MA15. I mean, I’ll see MA15. But, a line appears on the chart, which behaves quite differently and differs from MA15.

And, theoretically, this line on the chart will take into account current information about the latest prices for short periods, as well as the global trend, while not being sluggish. If such a non-DOM doodle carries up-to-date information, then it can certainly be used in some other way (maybe as a filter). i just want to twist it in my hands.

## filpan

In my opinion, your suggestion makes sense, when summing up taking into account the lag of each SMA, and keeping in mind:

forum on trading, automated trading systems and testing of trading strategies

Difference calculus, examples.

Aleksey Panfilov, 2020.06.22 21:18

I thought I should fix it in this branch. ))

forum on trading, automated trading systems and testing of trading strategies

Predicting the future using Fourier Transforms

Dmitrii, 2008.04.25 15:47

Oh, this Fourier…. And you know that the 0th harmonic is equal to the simple average (SMA) for the same decomposition period (T).

forum on trading, automated trading systems and testing of trading strategies

Predicting the future using Fourier Transforms

keeper, 2008.04.25 15:58

perfectly normal, it’s expected

It seems to correlate with:

forum on trading, automated trading systems and testing of trading strategies

Difference calculus, examples.

Aleksey Panfilov, 2018.01.31 09:38

SMA from SMA is, in my opinion, “not in the eyebrow, but in the eye” and by the way about Fourier. maybe we’ll get there again.

P/S 01.02.2018 I think there is also a difference, try to repeat your Banzai.mq4 indicator on this line.

SMA by nature is a saw only very small, so when taking the second difference, I assume a sawtooth line of the indicator.

Only this topic will not be easier than the topic from the previous post. i’m not sure i’ve quite figured it out myself. And in the subject of the previous message, everything is just transparent.

## nektomk

there will be something like an ema. Calculate the matrix of weights, and it will be. well, “lag behind” the price for about 60 bar.

the same old price, but for some reason crookedly interpolated. With large periods, sorry, everything is known about the past thoroughly – you can do better, the data is all there.

## capitalplus

Nicely

## khorosh

i strongly disagree. here today on the jew made 2 transactions. ma with a period of 200 and its levels are used for entry and exit.

## filpan

If suddenly, you asked about it.

forum on trading, automated trading systems and testing of trading strategies

On the lag of moving averages

Aleksey Panfilov, 2018.02.21 04:35

I like this image of the EMA. In my opinion, it is very clear and simple. We see the nature of the construction and, accordingly, the delay.

The lever of Archimedes, playable in the distance gain in weight.forum on trading, automated trading systems and testing of trading strategies

Difference calculus, examples.

Aleksey Panfilov, 2018.01.15 08:07

To begin a series of comparisons, let’s go back to the well-known one.

a1_Buffer[i] =

iMA(NULL,0,145,0,MODE_EMA,PRICE_OPEN,i);// a1_Buffer[i]= ((open[i] - Znach) +5061600*a1_Buffer[i+1 ]-7489800 *a1_Buffer[i+2 ]+4926624*a1_Buffer[i+3 ]-1215450*a1_Buffer[i+4 ])/1282975;a5_Buffer[i+92]=a1_Buffer[i]; if(i<=1) { for(z=92-1;z>=0;z--){ a5_Buffer[i+0+z]= 2*a5_Buffer[i+1+z] - 1*a5_Buffer[i+2+z] ; }}

a2_Buffer[i]= ( (open[i] - Znach) + 72 *a2_Buffer[i+1 ] )/73;

a6_Buffer[i+92]=a2_Buffer[i]; if(i<=1) { for(z=92-1;z>=0;z--){ a6_Buffer[i+0+z]= 2*a6_Buffer[i+1+z] - 1*a6_Buffer[i+2+z] ; }}

the first line

a1_bufferis built as a classic ema with a period of 145 (72*2+1), based on the opening points and shifted 72 steps to the left. A wide gray line in the drawing.The second line

a5_buffershows extrapolation using a straight line over the last two points of the first line. the gray line is thinner in the drawing.the third line

a2_bufferis constructed directly from the difference equation of the first degree. The blue line in the drawing.1*Y_(-1) –2*Y_0+1*Y_ (+1)=0 First degree difference equation for equidistant points.

2*Y_(-1) –3*Y_0+1*Y_ (+2)=0 Difference equation of the first degree for the shoulder in 2 intervals.

72*y_(-1) –73*y_0+1*y_ (+72)=0 first degree difference equation for the shoulder in 72 intervals.This is essentially an equation

Archimedes ‘ leverof the first degree.The fourth line

a6_buffershows extrapolation using a straight line over the last two points of the third line. The red line in the drawing that is based on the opening point.We see the complete identity of the two constructions. I gave an example so as not to show the conversion of the code and the form of the classical EMA formula to the difference form.

## capitalplus

I am curious that the indicator should take into account the average price not for one large period (where, as you say, the price will be monotonically growing), but for several different ones, but all of them would end at the current price. And the weight would be distributed from the smallest period to the largest in descending order. You should get an EMA that will move more actively, but at the same time take into account the latest trends.

Programmers and traders as soon as not perverted with mashki, but I can not find my own version.

## nektomk

do you even know how the EMA is considered, what it means, how it is used (in addition to forex nonsense) ?

it does not have to be considered as weighted-the incremental calculation is even faster than the sma.

for large periods, the (suddenly) monotonically growing weighted averages are interchangeable.

PS/ didn’t mean to offend, maybe harsh..But, ema more than 100 is a complete garbage, only for the joy of the eyes (it is visually smooth).