Elite indicators :) - page 395

[Mladen Rakic]

...

Here you go

regards

Mladen

clc4x:
hpt_-_wsro.mq4 Could you please make it to support MTF? Thanks

[umeshkathuria]

Hi mladen,

I'm trying to build this EA based on your "Sto_CCI_RSI 2" indicator given here.

I'm using this indicator on two different time frames but the EA is not considering the condition for higher time frame indicator.

Request you to kindly rectify this.

Thanks,

Umesh

mladen:
devinci

Here you go

There were some problems in the code that caused the repainting in the previous version (actually there was not limit to how many bars it could repaint since some variables states were inherited from previous calculation loop, and those were causing problems when new ticks were coming in - the longer you left it working, the bigger the distance in bars for that error could be)

It is corrected now and mtf and alerts are added (here is and example in a mtf.mode)

As of usage from an EA : the simplest way is the following :

int trendNow = iCustom(NULL,0,"Sto_CCI_RSI 2","",RsiPer,StochPer,CCIPer,Hot,Smoothing,4,0);

int trendPre = iCustom(NULL,0,"Sto_CCI_RSI 2","",RsiPer,StochPer,CCIPer,Hot,Smoothing,4,1);

if (trendNow!=trendPre)

if (trendNow==!) ... code for buying

else ... code for selling

regards

Mladen

[Mladen Rakic]

Umesh

Here you go

regards

Mladen

umeshkathuria:
Hi mladen,

I'm trying to build this EA based on your "Sto_CCI_RSI 2" indicator given here.

I'm using this indicator on two different time frames but the EA is not considering the condition for higher time frame indicator.

Request you to kindly rectify this.

Thanks,

Umesh

[umeshkathuria]

Thanks:)

mladen

As usual..Thanks a lot for the prompt reply:)

Umesh

mladen:
Umesh

Here you go

regards

Mladen

[clc4x]

mladen:
Here you go

regards

Mladen

Thank you so much. Please also have a look of my request on ""Trading the Trend" by Andrew Abraham" to exclude Sundar bar.

[camisa]

mrtools:
Kind of experiment that seems to work well so far, its a normalized jurik Rbci mtf with alerts, the Rbci was optimized for 1 hour and 4 hour eurusd.

does it recalculate past bars???

[san4x]

mrtools:
Kind of experiment that seems to work well so far, its a normalized jurik Rbci mtf with alerts, the Rbci was optimized for 1 hour and 4 hour eurusd.

Optimized using Digital Filter Generator?

Thanks, looks great by the way..!

San.

[Mladen Rakic]

...

Confidence bands of averages but a bit extended ...

---

Previous version was using a "simple" inverse cdf of a normal probability distribution for a confidence bands calculation. The obvious good about that manner of calculating is that it is simpler to calculate it than some other ways of finding out confidence (it will be obvious if you take a look at the code needed to calculate all needed in this one). The bad is that it neglects the degree of freedom (some more info about the degree of freedom can be found here : Degrees of freedom (statistics) - Wikipedia, the free encyclopedia )

This version uses different calculation for confidence bands (Student's T) which takes into account the degree of freedom of the average too. Some more information about Student's T and how it can be used for confidence bands calculation can be found here : Student's t-distribution - Wikipedia, the free encyclopedia Regarding degree of freedom one assumption is made even though it is not always the truth : it is set to length-1. For example linear regression should be length-2, but I let it here as is since linear regression is yet another cookie that will be dealt with separately since the "2" in the degree of freedom gives us more opportunities to examine (and study) more and more precisely

If this one is compared to the previous than it will be wider for shorter periods and it will be narrower for longer periods unlike the previous version which kept the "width" linearly. It is natural to have the "adaptive way" *depending on length) since the longer the sample is the more confidence we can have that it is OK (remember that here we are talking about confidence bands of averages). Other than that, indicator behaves similarly as the one before (with mtf already in it) Again it seems that shifting confidence bands can give us signals (especially with a bit lower (but not too low) percentage for confidence bands - 75% on lower 30 period NonLag MA example)

PS: made some changes in code (it affects bands with confidence levels less than 50%). Please re-download the indicator if you do not have the latest version

[William Snyder]

camisa:
does it recalculate past bars???

Hi Camsa,

No recalculating on closed bar.

[William Snyder]

Snowski:
Optimized using Digital Filter Generator?

Thanks, looks great by the way..!

San.

Hi San,

Yes used the digital generator using Alpari on mainly 4 hour data!