Indicators: Smoothed Accelerator
Для компенсации задержки проходы производятся навстречу друг другу.
- Please clarify on delay compensation.
- Is it redrawing from here?
- Please clarify on delay compensation.
- Is this where the redrawing comes from?
If even the most common MA is run through the array first in one direction and then in reverse, the delays of the counter passes are compensated and the result is a curve that has no lag from the original sequence. Unfortunately, this will result in redrawing at the edges of the sequence. How much the tail will wag depends on many factors, but it will definitely wag with this approach.
Yeah, that's what I figured, thanks.
P.S. I came across your articles. Reading.
If even the most common MA is run through the array first in one direction and then in reverse, the delays of the counter passes are compensated and the result is a curve that has no lag from the original sequence. Unfortunately, this will result in redrawing at the edges of the sequence. How much the tail will wag depends on many factors, but it will definitely wag with this approach.
How is it that the delay is compensated in a miraculous way?
Do you think that just because you run the mashka back there, you'll know the future?
ZY the most common fly, as you put it, has a delay of half a period, so wherever you run it, the delay will remain in place.
How does that miraculously compensate for the delay?
do you think that just because you run the mashka back there, you will start to know the future?
ZY the most common wizard, as you put it, has a delay of half a period, so wherever you run it, the delay will remain in place.
Let's make an experiment:
- Copy the original sequence of sufficient length into the working array.
- Smooth this data in the array, for example, using MA2.
- Change the indexing direction of our working array.
- Let's perform MA2 smoothing again
As a result, we will get the smoothed original sequence located in our working array. The result of smoothing will correspond to the application of a zero-delay filter (symmetrical impulse response) except for the sequence edges where the edge effect or as it is called here, redrawing will appear.
Above, MA, i.e. a filter with a finite impulse response, was used as an example. When using filters with infinite impulse response (e.g. EMA), theoretically the edge effects will propagate over the entire length of the sequence. But if the smoothing factor is small enough, this feature can most often be neglected.
If you cannot theoretically understand why this happens, perform an experiment, smoothing, for example, a single pulse located in the middle of the sequence, and make sure that there is no delay after filtering. By the way, instead of multi-pass algorithms you can apply smoothing once with a filter having a symmetric impulse response. The result will be the same.
As it was already said, at counter-passes of smoothing, the filter delay is compensated, and it happens objectively and does not depend on how you treat it.
PS
It should be noted that edge effects (redrawing) are determined by the resulting impulse response of the filter and the shape of the original signal. In other words, the redrawing does not occur chaotically, but according to a strictly defined law, which allows you to use the redrawing indicator for making various kinds of decisions. If, of course, the algorithm of making these decisions is not too primitive.
good indicator, but it doesnt show up in some charts
Hello! The indicator is WANTED!) Just a question: Why doesn't the indicator work on the Dow index? On HQ and ES works great! But it doesn't work on Dow((( Help!
Works on Dow
Dear Victor,
I am trying your indicator and I have to say that it gives some insight in advance. Obivisly it is necessary to use them understanding the global context of the price action. Particularly it is interesting the inversion point of te indicators above all when there are, before and after them, steep slopes. But I have not understood the equation that you are using and why you are calling it acceleration. You have a division operation between two close prices; my expectation was instead a double "derivate" with respect to the time something like
- V(i) = ( P(i) - P(i-n) ) / TF*n, where P is price at the i-th candle n is period and TF the used timeframe
- A(i) = ( V(i) - V(i-n) ) / TF*n
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Smoothed Accelerator:
The indicator plots the smoothed value of the Accelerator.
The smoothing is carried out by four-pass EMA with period, specified in the "MA period" input parameter.
Author: victorg