Requests & Ideas - page 75

 

Odysseus

Will check it and will get back with what I discover

regards

Mladen

odysseus11:
Mladen,

I turn to you for help on this one!

Mr. Tools was kind (and smart) enough to attempt to add newsfilters to the 2 EAs I have attached (polyfit scalper 1.59 and tickscalper 3.44)

This is great in theory, but I have observed several instances in the past couple days where the text on the screen will say "Avoiding high impact news" and yet the EA will open a trade nonetheless (usually with horrible results!), so apparently the newsfilter doesnt work properly. It uses the TSD news 1.5 calendar. I have attached that zip as well. Could you PLEASE take a look at the EA(s) and confirm that the news filter does in fact FILTER the entries during the correct news times? Also, is there any need to in any way adjust the specified time zone in the EA to match the calendar time zone? (I dont think this is the problem because the text comments *do* say "avoiding high impact news" at the right times, I think the problem is in the filtering of entries because entries are STILL occuring when they shouldnt be).

Thanks so much, I asked Mr. Tools, but he didnt see any issues so you are the one who can help --- very much appreciated!

Odysseus
 

MTF G browser

I know this has been touched upon before (and has raised be quite a lot of interest), but as far as I am aware we still do not have a MTF G browser (with all due respect, Mr tools' version is not ideal).

A kind request for Mladen - I know you have attempted this before, but wondered if you could please have another look at the G browser and make it MTF. What do you think? Thanks in advance.

FL

 
 
 

Was trying to find an appropriate smiley but could not find one, so just the usual one at the start Now you are going to have 2 versions (thanks mrtools )

_______________________

simba this is the modified version you already are partially aware of : it has an addition of choosable sample size. The default is max period * 3, but the truth is that it is a minimal bars required for Goertzel calculation. In some cases increasing sample size compared to the minimum can give very, very different results and it seems that has to be tested thoroughly : simple logic s saying that the larger the sample, the more reliable should be the result, but the question is "how reliable we want the whole thing to be?" I know that the question sounds silly, but think of it (especially when it comes to "prediction" : how reliable do we want the prediction to be (or at least to think of how reliable it is)?).

The rest is more or less the same as in the previous version (some code changes were done, but they are not too significant, the biggest changes are the sample size addition and degree display of phase instead of radian display in the phase part) Time frame can be chosen but when in mtf mode it will not show the predicting part (as I have mentioned once - mtf-ing the predicted part is a bit of a problem when missing bars happen on a chart and then some ridiculous results can happen, that is why there is no predicted part in the mtf mode - metatrader does not have, like some other platforms a possibility to draw values in the "future", all it knows is to draw in the past, and when past bars are messed up due to brokers playing their games, the future is messed up too - no perfect future can be achieved in metatrader, sort of speaking ). List of cycles displayed is a list of cycles of target time frame so the list of cycles is in mtf mode too (it is showing proper information too)

Here is an example of a 1 hour Goertzel browser on a 30 minute chart with sample size set to 1000 and 10 top cycles used for generic cycle construction (so it is using 3 times more past bars than the default in this example)

when it is at 1 hour chart, then the "prediction" is visible too

Files:
 
mladen:
Was trying to find an appropriate smiley but could not find one, so just the usual one at the start Now you are going to have 2 versions (thanks mrtools )

_______________________

simba this is the modified version you already are partially aware of : it has an addition of choosable sample size. The default is max period * 3, but the truth is that it is a minimal bars required for Goertzel calculation. In some cases increasing sample size compared to the minimum can give very, very different results and it seems that has to be tested thoroughly : simple logic s saying that the larger the sample, the more reliable should be the result, but the question is "how reliable we want the whole thing to be?" I know that the question sounds silly, but think of it (especially when it comes to "prediction" : how reliable do we want the prediction to be (or at least to think of how reliable it is)?).

The rest is more or less the same as in the previous version (some code changes were done, but they are not too significant, the biggest changes are the sample size addition and degree display of phase instead of radian display in the phase part) Time frame can be chosen but when in mtf mode it will not show the predicting part (as I have mentioned once - mtf-ing the predicted part is a bit of a problem when missing bars happen on a chart and then some ridiculous results can happen, that is why there is no predicted part in the mtf mode - metatrader does not have, like some other platforms a possibility to draw values in the "future", all it knows is to draw in the past, and when past bars are messed up due to brokers playing their games, the future is messed up too - no perfect future can be achieved in metatrader, sort of speaking ). List of cycles displayed is a list of cycles of target time frame so the list of cycles is in mtf mode too (it is showing proper information too)

Here is an example of a 1 hour Goertzel browser on a 30 minute chart with sample size set to 1000 and 10 top cycles used for generic cycle construction (so it is using 3 times more past bars than the default in this example)

when it is at 1 hour chart, then the "prediction" is visible too

Mladen and Tools,

Thanks to both for your versions,I think that now Advanced Elite Goertzel Browser is state of the art...Fantastic job, both of you.

 

mrtools ; Mladen

One could modify AutoFib TradeZones to see on the right with all the% (0%, 23.6%, 38.2%, 50%, 61.8%, 76.4%, 100% ... etc), the cash value each. .. %, of the diviza?

Files:
 

Price Targets

What indicator is best suited for a logical exit point - Stop and Limit?

These are the ones I've used so far...

Murray math - tricky!

Fib levels - not sure if its reliable, plus too much clutter on the charts

On my will - has not always been wise

Any ideas, thoughts or words of wisdom? What do you fellow traders use for a succesful and logical exit

 

Values of Elliot Oscillator colored bars

Hello mladen

A feature request to your indicator "Elliot oscillator - waves 1.01"

Is it possible to label values for the colored bars?

For example, in the attached picture, the value of the first yellow bar (Elliot oscillator 5,35) is 0.0065. Can you make this value to be printed as 65 (*10000) on top of the colored bar? (On the top for yellow bars and below purple bars)

The value need to be printed ONLY for the colored bars and stays in history as long as the bar remains colored - meaning, if a new colored bar is formed replacing the previous one, the old bar should no longer have the printed value.

Thanks a lot in advance. The Elliot oscillator has a amazing potential!

Files:
elliot_1.gif  18 kb
 

Phuh

I hate when I can not find a solution, and not having a solution for mtf-ing extrapolated values in Goertzel browser was one of those things. Well we have now a mtf of a future value too

All it took is to forget the previous way of thinking and think in a completely different direction. If metatrader does not have a perfect future bars, than we should make our own. And this is the results (the best thing is that all the drawing of prediction is done by the target time frame - nothing is left to chart time frame to do )
One detail maybe needs explanation : the displacement that is visible while the bar is still not closed (target time frame bar) I left it that way at purpose : Goertzel value at that bar is exactly as it is show, predicting value is showing what will it be if it does not change, so even the current bar is taken as a prediction in this case (this is the case I am talking about)
PS: parameter FutureLineDisplaceis for internal usage only - whatever value you enter, it will be ignored. It is used as a mean of communicating some data from one time frame to another
Reason: