The JMA of the Allsto is not so beautiful as the STO JMA of Spiggy
As you see STO JMA give very good entries/exit (parameters need to be optimized) when we follow the trend (supertrend, fractal, take a higher arrow in a uptrend, lower arrow in a downtrend etc...)
Try out this version. It seems that I have found what was going on but it still needs to be tested more thoroughly. As far as I see there were cases when simply because there was more than one signal on the same bar (and in it there can be up to 3 arrows on the same bar), a new arrow drawing was causing that the previous arrow disappear
So beautiful (with the 1.03 version of Mr tools)...
It test it immediately
Ps: Euh when do you sleep ?, it's 11:18 AM in France
We are in the same time zone
Sorry but some problems persist with this 1.04. (Graph DAX ut1) (but it seems to have less problem than before )
As you see on my graph, it seems there is a conflict between the arrows
The arrow when %K cross % D in the OB/OS signal doesn't appear when there is another arrow (%K cross the OB/OS signal for example)
Thanks a lot
Ps: I thought you were american
As a simple test of what that JMA is doing, do the following : change the BarCount from 900 to some other number and see the arrows but also see the values of the indicator itself (I changed it to 1500). Then change it back again. It means that that JMA is 101% sure repainting and that it largely depends on the bars that are used for calculation (repainting will always happen when a new bar is opened and it will repaint BarCount back - so no limit of how many bars it repaints back)
That is a well known problem with almost all JMA calculations (even the "Kositsin" version) and that is why I, for example, prefer some other indicators to JMA
PS: mrtools is American I am not
Ok I try
You were right Master !
See the difference with 900 and 1500
I don't know who it can be useful to repaint many many bars before, pffff...
Do you think it can be solve or I just need to cry
Do you try do put the signals only on the bar close ?
As I said : JMA is like that. Nice but does what it does. It is happening because it inherits some previous bars values in a way that even the last bars value depends on a first calculated bar. Try using large number for BarCount (5000 for example) but even that will not prevent things like that with JMA (as a simple proof : change the BarCount from 5000 to 5001 and see how the value of stochastic changes - that is what I am talking all the time about - the repainting, since with 5000 bars 1 bar difference must not cause difference in values)
Thanks a lot mladen,
It's what I tried
In visual/test mode with 200 bar count there are a lot of repaint arrows
With 1500 not any arrow move
I let the TG analyser for 30 minutes and I will see (You have 30 minutes to breathe )
What a wonderful indy this TG analyser (thanks Mr tools)
"Bon appétit" as we say in France