• Get access

Requests & Ideas - page 104

Mladen Rakic
163289
Mladen Rakic  

Caprica

In general, the longer the maxper, the more cycles will be identified and the extra cycles identified are probably going to be longer - now I am talking about "regular" cycles search, not the detrended cycles search. To do that turn the DetrendSmoothData to false. With detrended data it is a completely different game and than the main "work" is done in the Hodrick Prescot filter so you should concentrate on HP parameters then.

So, in "regular" mode, it depends what are you after : if you are after long cycles use longer maxper, if you are after shorter cycles, use shorter maxper.

regards

Mladen

caprica:
Hi Mladen and Pip

Can I ask what your recommended settings are for maxper? the more I play around with this setting the more I get concerned about the ability for this indicator to find meaningful cycles.

thanks, Caprica
Flytox
256
Flytox  

Thank you very much brax, this is perfect.

brax64:
Hi Flytox,

I did what you asked for, now it's possible to choose in which window you want the info panel and with the parameters Pan_X and Pan_Y you can pinpoint the position.

In the meantime, I've add also the possibility to calculate retracement and extension, for that now is version 6.5

This thing is going to be really addictive ...

Please read the instruction in the code and let me know if it match your expectation.

Best Regards
san4x
324
san4x  

Simple Idea

Hi NewDigital,

Perhaps a handy idea:

As this section is growing nicely, have the Advanced Elite section organized similar to Elite section (Indicators, EAs, etc).

Cheers,

San.

Sergey Golubev
Moderator
114836
Sergey Golubev  

yes ... may be ...

but advanced elite section is small one compare with simple elite.

And there are not so many advanced elite EA for now.

Ricon
54
Ricon  

Immediate Entry & Immediate Trailing EA

Apologies if this has already been asked somewhere, but this site is so large i could spend weeks reading it!

I'm looking for an EA with the following characteristics:

1) Enters trades immediately, along the lines of Alejandro Galino's Multi Lot Scalper (although i'm not really interested in multi lots). Not randomly, but the criteria for entering trades isn't complicated eg maybe whichever way an MA is going.

2) Has a trailing stop feature that allows you to specifiy when to start trailing eg if i have a TS of 40 and i specify a start trail of 10, that means the EA will begin trailing at 10 pips profit, not wait until it reaches 40 pips profit. So if i have a SL of -60, and my trail starts at 10 with a TS of 40, when i reach 10 pips of profit my SL moves to -30. EDIT: I realise this isn't immediate trailing like i mentioned in the title of this post, but i made a mistake there and don't think i can change the title now.

3) Has no restrictions on timeframe.

4) Has a maximum spread allowed parameter. This is almost standard these days, so i don't expect it's a big ask.

5) Allows you to specify a TP and SL.

6) Has money management.

Hope someone can help!

brax64
158
brax64  

smoothed klinger oscillator...

Hi Mladen,

I was wondering if could be possible to convert this indicator XKVO - MQL5 Code Base from mq5 to mq4.

Thanks in advance if you will find the time.

Best Regards

Mladen Rakic
163289
Mladen Rakic  

brax

Here you go You will notice that it uses only non-adaptive indicators (the ones that I call a "single parameter" averages) and I would like to explain why (it will take some time though ).


First I will start from metatrader 5 and some comparison : on this picture the upper macd is the macd of volumes (xkvol indicator) using "jurik" smoothing. And ,on a first glance, everything is just fine.

Well, it would be if there was not that other instance of the same indicator using T3 instead. And there comes the problem : "Jurik" moving average should not give results that are an imitation of T3 (even more on the example, T3 is faster than "Jurik" example !!!). Jurik moving average is an adaptive moving average T3 is not. The fact that the results are so similar shows just one thing : that something is wrong with that "Jurik" smoothing (it is not adapting at all, for one). On a long calculating periods adaptive moving average should always be much faster on sudden smoothing values change from a non-adaptive one, and, in this example the changes are as fast as it gets (from one bar to another it changes from +x to -x (and vice versa) without any intermediate step at all) and thus difference should be big.

And that speed makes adaptive averages "not good" for macd calculation : the nature of macd is that it counts on the lag of averages in order to avoid whipsaws and to calculate a "trend" having a lag counted in too. If lag is avoided, macd will not give reasonable trends (try it out with kama and ama (which are known for a long time and there is no coding mystery in them) in xrvol indicator and you will see what I mean - those result seem meaningless until you remember that kama and ama are adaptive).

So, if the "jurik" is coded as it should, than it should look something like this

As you can see, a proper adaptive indicator again gives meaningless macd in this case of very fast "volumes" sign change. Since I am not going to seek what is wrong with Kositsin's "jurik' (he tends to write overcomplicated code, and in metatrader 5 he does it even more - I think he should revise his code not me), I decided to leave jurik out of this. Added T3 (since it is a non-adaptive as I explained and since it gives interesting results as far as this indicator is concerned) The rest of possible moving averages is our standard 17 types of averages (not going to list them again in this already lengthy post) Just attaching an example of T3 version (MaMode set to 17)

Hope you don't mind the lengthy post

regards

Mladen

brax64:
Hi Mladen,

I was wondering if could be possible to convert this indicator XKVO - MQL5 Code Base from mq5 to mq4.

Thanks in advance if you will find the time.

Best Regards
bebeshel
519
bebeshel  

macd of volume

You could use your "Jurik filter simple 1", and not Jurik made by Kositsin's.

jurik_filter_simple_1.mq4

brax64
158
brax64  

Hi Mladen,

Thanks a lot for your always priceless explanations, I really like to hear your point of view; constantly supported by sound evidences!! So no worries about " lengthy post..."

Again your job here for the all community it's simply amazing, my deep appreciation and gratitude to you!

Best Regards and ... happy Halloween !!

mladen:
brax

Here you go You will notice that it uses only non-adaptive indicators (the ones that I call a "single parameter" averages) and I would like to explain why (it will take some time though ).


.......

......

Hope you don't mind the lengthy post

regards

Mladen
William Snyder
9493
William Snyder  
bebeshel:
You could use your "Jurik filter simple 1", and not Jurik made by Kositsin's. jurik_filter_simple_1.mq4

Please read Mladen's post again. To make it easier, and please understand Jurik filter simple 1 is adaptive like he said below :::::

And that speed makes adaptive averages "not good" for macd calculation : the nature of macd is that it counts on the lag of averages in order to avoid whipsaws and to calculate a "trend" having a lag counted in too. If lag is avoided, macd will not give reasonable trends (try it out with kama and ama (which are known for a long time and there is no coding mystery in them) in xrvol indicator and you will see what I mean - those result seem meaningless until you remember that kama and ama are adaptive).