Dynamic zone indicators ... - page 113
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
yes.... cci_adaptive_smoother.mq4
It is a nice one
yes... cci_adaptive_smoother.mq4
Here is a bit upgraded (more user friendly) version of that indicator : cci_adaptive_smoother_1_1.mq4
Here is a bit upgraded (more user friendly) version of that indicator : cci_adaptive_smoother_1_1.mq4
dear mladen, do we have " Dynamic zone - pa fisher transform (or inverse fisher transform) of CCI " already?..... i'm not able to find one....
dear mladen, do we have " Dynamic zone - pa fisher transform (or inverse fisher transform) of CCI " already?..... i'm not able to find one....
as to my knowledge,pa ift,yes......of cci,no
I'm using DZ algorithm for both MT4 and MATLAB and I noticed some discrepancies. I applied DZ bands to TSI both in MATLAB and in MT4 for the same time period. For MATLAB screen shows a collection of one minute results (so indicator is calculated ever minute but only last minute data is collected), for TSI its done in normal way with 'indicatorcounted' and dll from 2010. TSI values seems to be the same (see value 5.96 on both screens) but values of bands differ in areas when band is changing the direction so it looks that at least for MATLAB this algo is unstable. Code of the algo for MATLAB is like shorter code for Dynamic zone MACD from 1st posts. Any idea what it can be a reason of this instability ??
I also noticed that there are two versions of this code: shorter in Dynamic zone pricezone1 and longer in Dynamic zone pricezone. Any idea whats a difference between them ??
Krzysztof
I'm using DZ algorithm for both MT4 and MATLAB and I noticed some discrepancies. I applied DZ bands to TSI both in MATLAB and in MT4 for the same time period. For MATLAB screen shows a collection of one minute results (so indicator is calculated ever minute but only last minute data is collected), for TSI its done in normal way with 'indicatorcounted' and dll from 2010. TSI values seems to be the same (see value 5.96 on both screens) but values of bands differ in areas when band is changing the direction so it looks that at least for MATLAB this algo is unstable. Code of the algo for MATLAB is like shorter code for Dynamic zone MACD from 1st posts. Any idea what it can be a reason of this instability ??
I also noticed that there are two versions of this code: shorter in Dynamic zone pricezone1 and longer in Dynamic zone pricezone. Any idea whats a difference between them ??
KrzysztofKrzysztof
The "longer" version of dll is optimized and faster - and is in some parts, differing from the mql code (results wise too - that is done in order to optimize speed and precision). That version should be used (for the sake of speed, if not for anything else, but the main reason for changes was to add precision, not speed)
______________________________-
As of discrepancies : if the data is the same. the results must be the same (I am assuming you mean when dll is used from metatrader and matlab). That can be the only reason for difference. No idea how the data was fed from matlab, but there should be no instability if the data is fed properly to the dll. If, on the other hand, matlab code is used for dz calculations, then the reason of difference should be obvious - the "basic" dz algo (the one from the 1sr post) lacks precision when very small or very large values are used, and that code should be altered to correct that - that is what was done in the newest ("longer") dll
Hi,
I was using this version of DZ code in MATLAB script (not dll) and dll from 2010 in MT4 (e.g for DZSell)
if (prob<(initValue-eps))
{
right = yval;
yval = (yval+left)/2.0;
}
else
{
left = yval;
yval = (yval+right)/2.0;
}
delta=yval-left;
[/CODE]
however later you were using this one
[CODE]double prob = count/lookBackBars;
if (prob>(initValue+eps))
{
left = yval;
yval = (yval+right)/2.0;
}
if (prob<(initValue-eps))
{
right = yval;
yval = (yval+left)/2.0;
}
if (prob(initValue-eps))
{
left = yval;
yval = (yval+right)/2.0;
}
delta=yval-left;so maybe this is a reason of discrepancy as dll has the 2nd version or ?? Probability was set to 0.05, LookbackBars = 900. So you think if I will change to longer version inside script it will match or improve ??
Krzysztof
Hi,
I was using this version of DZ code in MATLAB script (not dll) and dll from 2010 in MT4 (e.g for DZSell)
if (prob<(initValue-eps))
{
right = yval;
yval = (yval+left)/2.0;
}
else
{
left = yval;
yval = (yval+right)/2.0;
}
delta=yval-left;
[/CODE]
however later you were using this one
[CODE]double prob = count/lookBackBars;
if (prob>(initValue+eps))
{
left = yval;
yval = (yval+right)/2.0;
}
if (prob<(initValue-eps))
{
right = yval;
yval = (yval+left)/2.0;
}
if (prob(initValue-eps))
{
left = yval;
yval = (yval+right)/2.0;
}
delta=yval-left;so maybe this is a reason of discrepancy as dll has the 2nd version or ?? Probability was set to 0.05, LookbackBars = 900. So you think if I will change to longer version inside script it will match or improve ??
KrzysztofKrzysztof
The dz algo that you are using (and that is the original dz algo) depends very much on the initial value and number of steps. When values out of the expected range (and originally dz was intended to be used on rsi and stochastic like indicators - with bounds very well known) are used as a source for calculation, it loses it precision and can lead to unexpected results. That is why I suggested that dz algo should be modified (and that was done in that longer version of dll)
Added dynamic zones to the version updated by Mladen.
dynamic_zones_of_cci_adaptive_smoother.mq4
Added dynamic zones to the version updated by Mladen.
dynamic_zones_of_cci_adaptive_smoother.mq4
Dearest MRTOOLS,
so much thanks for the DZ adaptive smoother ver of cci,so nice of you,can you add center line of zones too,as in mostly DZ indicators.thanks.
regards