Yes It is a "variation" on that one but moved the smoothing from "post-calculation" to "pre-calculation" (I really was curios what would it produce that way) Adding the rest of averages was just a + to be able to test pre-smoothing with the usual averages we have
As I understand - it is some kind of variation of TTM/Trend scalp indicator/system from here, right?
I remember - I traded it with good results here:
If yes so thanks a lot.
I updated this link:
How To Trade
PS: I usually use Kalman filter on 1 hour or 4 hour chart (something like this : this is a 4 hour EURUSD chart. I recommend you look at it in the past too since there are some very, very interesting parts showing how the filter behaves in some periods
Please could explain a bit more about how you trade with this?
Look at the cross of the central lines (the two Kalman filters) as well as the cross of the prices over the central lines. Bands are there for eventual targets and stop losses. It is a discretional way of using it to find the trends out and normally I am the one that decided which time frame to use. As I said, when I want to see the "main" trend I go to 4 hour time frame and when I want to determine entry or exit points than I go to 1 hour time frame keeping in mind what 4 hour time frame shows.
usually you will see that when the price crosses central line and retraces, Kalman filter lines did not (and it is not due to lag of them). Sometimes, when I am a bit more "freeminded", it looks like the filter has a mind of its own. Takes some time to get used to it, but that is the one I am using as a basis of my trading and I like its "behavior". Just to remind what some people know : I trader only EURUSD so this template is more or less tailored to it
Could you fix the arrows & alerts & could you create the trend buffer?
Snake (that is used in that indicator) recalculates (it is in fact a centered TMA) which means that signals will change places as the "snake" changes values. So, if buy "fixing" you meant to keep them on same place (not repaint) using "snake" it is not possible. Let me know if that is what you meant or you have some other idea
I would rather "no repaint" if possible. What would you suggest to replace the Snake?
Thanks for the explaination.
As far as results are concerned, centered moving averages are a class of their own. Will try with some other averages to see if some will give comparable results, but can not promise anything too similar to what centered TMA produces
Is there any way to tell or control how far back the TMA will re-paint. Fractals are suppose to re-paint 2 back, Henderson is said to be 2 back but I've seen it change 4 back. I seem to be getting more involved with TMA's. I would like to be able to spot there vulnerability/chart delay time, for instance this TMA being used in 10.2, attached. How far back will it change?
cenetered TMA repaint exactly half period bars back. In the case of the indicator you attached it is with default setttings 20
Been thinking about this problem and here is my conclusion :
Combining 2 types of averages crosses actually can not be automatized. The problems is that we are comparing apples with pears. In cases like MACD we are having same type of average and in that case it is simple and quite obvious which one is faster : the one with shorter calculating period. In case of comparing 2 types of averages that criteria is completely lost : the code can not "know" which one is the faster one (just take an example of SMA and NonLagMa : non lag ma is going to be faster even for longer periods of calculations)
And since it does not "know" the faster and the slower, a 2 type averages cross indicator also can not know the trend (for the reason explained). Trend can be found only in cases when the types of the averages are the same. I can mark that "verage 1" is over or under "average 2" but in terms of "trend", when averages are of different types, it will mean nothing in some cases