EA Switcher - page 8

 
dswk:
Yep....

Hi everybody, i'm back.

Slowly recovering, yes NewDigital ...lol... but back in the market... and now wake-up this thread....

Best wishes for 2007.

I will try to descibe some task for coder.

We need just the following:

- name the functions (just name only);

and

- connect all the functions in some simple switcher. Just to connect between each other.

All the functions will be developed in separated files.

 

Good to hear

Well, I guess I can take this as a suggestion to relax a bit and enjoy other things in life other than trading and planning to trade

Didn't mean to come across as abrupt - Sorry everyone

Doug

 
dcraig:
Well, I guess I can take this as a suggestion to relax a bit and enjoy other things in life other than trading and planning to trade

Didn't mean to come across as abrupt - Sorry everyone

Doug

I am not starting because I am not a coder. I will not code it. I am waiting for some reply from Igorad or some other coders. All that I can do is to define some tasks about what to do and how to do it in very easy going way.

 

This is an amazing idea!!!

newdigital:
Yes, the pairs and timeframes may be mentioned inside the code.

BTW I have some ideas about how to create Portfolio Manager. Steps of the development aee the following.

- step 1.

Create EA or tool which will automatically detect the condition of the market sending/saving all the information in txt file for H4, D1 and W1 timeframes.

I did it many times with Ichimoku system and with digital filters' system. Those systems are having great forecasting possibility and I used it and posted in many thread already. If I could do so EA can do as well. Because it was not something personal: I was reading the indicators only in fully programming way.

We may use Ichimoku chart to determing the market condition and do some forecasting. I did it here for example. Thread is here and here, some template is here. Some examples of estimation we may find in elite asctrend thread. Explanation about some indicators is here.

Besides we may use digital filters to do the same. I did it many times.

Just some usefull likns:

trading system thread, generator of digital filters, indicators' description, download digital filters indicators.

We may code some very preliminary templated EA and then, later on, insert all the indicators and systems (optional) to do the same (trading recommendation, market condition estimation, forecasting and so on) what I did but utomatically with txt files, email, and so on.

We may use the other systems (anyone) for market condition explanation (as UseIchimoku =true or false; UseDigitalFilters = true or false; and many other systems people suggest). Because I can do it manually with Ichimoku and digital filters' system, and the other people can do it with some other systems so we may create just a templated Portfolio Manager to insert all the systems inside (with text for txt file for all the cases of the market situation). The explanation/trading recommendations will be saved in txt file (it will be the same explanation as brokers are doing weekly, you somebody explaining on TV for example). But fully automated.

- step 2. Improve this Portfolio Manager. This EA will predict/forecast of the price movement with support/resistance and so on.

- step 3. EA will manage all other EAs in MetaTrader/charts switching off or switching on according to the market condition and forecasting. For example, if trend market - Portfolio Manager will switch on some EAs which are trend followers, if flat - scalpers will be switching on.

- step4. Portfolio Manager will use one Money Management for all EAs working in Metatrader.

- step 5. Portfolio Manager will re-act on the news. We should create or use some news calender and some tools or dll files to work with news.

- step 6. Portfolio Manager will generate trading signals and send them by email, sms and so on.

Step 1, 2, 3 and 4 is not very difficult. More difficult is step 5 and 6.

It may be the plan of our development.

Hello NewDigital & Dswk,

Guys, this is something that has come to my own mind. Since I have been reading all the trading EA threads on this forum (& other FX forums) it has become VERY clear that no one EA can fit ALL market conditions. Thank you for starting this thread.

I think organization of information is going to be key to success - may we should follow newdigital's catagories above to organize the information. So when someone post input, it should say which catagory they are providing information about.

I hope to be an active member of this thread.

Thanks

 

Changing Tactics From Choppy to Trend

I have an EA that I use mostly in a 5M time frame on EURUSD. It is very profitable when the market is choppy on the daily bars. The entries are counter-trend, so there are stops that get me out in an extended trend. The current strategy is profitable over the long term, even with the stopouts from the big trends.

However, I want to improve it with a trigger that recognizes an established trend and reverses the signals when that happens. There is a parameter in the code now that reverses the signals, but that has to be done manually.

The difficult thing about that, though, is knowing when to revert back to the choppy market signals. I am not a programmer; I'm just looking for ideas. Thank you.

 

Hi,

We still working on this kind of idea... not finished.

We pre-called it EA-Switcher : please have a look to : https://www.mql5.com/en/forum/175657

Our idea is to detect what kind of market is on the way, to choose the right EA to use...

Comments welcome.

 
dswk:
Hi,

We still working on this kind of idea... not finished.

We pre-called it EA-Switcher : please have a look to : https://www.mql5.com/en/forum/175657

Our idea is to detect what kind of market is on the way, to choose the right EA to use...

Comments welcome.

Yes, it is EA-Switcher. Still under development.

Because we was switched on to NewsTrader EA and after that we will be dropped to this EA-Switcher.

 

Simpler Idea

I'm thinking more simply than what I see posted throughout this thread. I have spent more than 20 years trying to figure out when the market has started trending in my time frame, and I haven't succeeded. So, the notion of using a bunch of traditional indicators for that purpose leaves me a bit cold. However, I probably don't have the brainpower of the folks I see posting here.

In the meantime, I have developed a system that is near flawless in a choppy market (as I said earlier). It allows a maximum number of entries; when I'm stopped out of them all, we're in a trend for sure. Therefore, my system identifies the trend when I lose; and that trend almost always continues for profit opportunities in that direction.

This suggests to me that the guts of two systems could be used to identify what's going on, if you can find two that are dependable in each kind of market. When the chop systems starts losing (on paper) and the trend system starts winning (on paper), the switch is made.

Just a thought for a slightly different approach.

 
newdigital:
I will try to descibe some task for coder.

We need just the following:

- name the functions (just name only);

and

- connect all the functions in some simple switcher. Just to connect between each other.

All the functions will be developed in separated files.

Why not making a master EA, that writes the name of the EA that should run to a file or writes a global variable with that information.

The other EA's (the ones which do the real trading) start with a function that looks at the file or at the global variable to see if it has to run. If not, it doesn't do anything.

So that would mean writing one master EA and one function (something like bool ShouldThisEARun) that could be added to the other EA's.

 
JosTheelen:
Why not making a master EA, that writes the name of the EA that should run to a file or writes a global variable with that information.

The other EA's (the ones which do the real trading) start with a function that looks at the file or at the global variable to see if it has to run. If not, it doesn't do anything.

So that would mean writing one master EA and one function (something like bool ShouldThisEARun) that could be added to the other EA's.

Interesting Concept ! Thanks Jos.

Reason: