Australian (ASX) Stock Market Forum

Forex Robots - legit or scam?

just above the green 'not accurate to last bar' writing, u should get a 'resize' window cursor which u can pull up to just under the statistics at the top, to give the graph more size.
 
yeah in the indicator properties (right click chart, indicators, select, properties).

also turn off all the prices and writing on the chart to get rid of the writing at the top left and the currency prices.

u can then pull the graph up using a resizing cursor to just below the indicator's green writing.

also u can adjust the time on the chart, which will show u different time periods on the EA.

The time view is tricky to handle. But wow, that looks ok.
 

Attachments

  • aea.jpg
    aea.jpg
    63.2 KB · Views: 5
Oh that last charts was with a starting capital of $828.00 USD. This following chart shows the starting capital of 475.00 USD that I actually started with originally. And that is after trading live for around 3 weeks
 

Attachments

  • aea.jpg
    aea.jpg
    67 KB · Views: 6
Well, I've taken a few steps forward. I'm finding it a bit overwhelming to learn MQL4, but in the meantime I've been learning VBA and have setup a spreadsheet to autotrade FX through IB. Have to wait till Monday now though before I can start testing it and debugging. It's not as simple as it seems to me as a programming beginner to encode human logic and have the PC check for all the possible conditional decisions needed to cover all bases. I'll get there though and am looking forward to this next stage in my trading journey.

I can now better appreciate those of you who have written EAs or other programs from scratch. Kudos to you. :)

Ps. Question for VBA programmers. In terms of processing usage and efficiency is it better to use more or less modules? As a complete noob, I'm finding it simple to write lots of little modules that do specific tasks and then write an overlay kind of module that calls the little ones as needed - but I'm not sure if that taxes the PC's processing too much?!
 
Well, I've taken a few steps forward. I'm finding it a bit overwhelming to learn MQL4, but in the meantime I've been learning VBA and have setup a spreadsheet to autotrade FX through IB. Have to wait till Monday now though before I can start testing it and debugging. It's not as simple as it seems to me as a programming beginner to encode human logic and have the PC check for all the possible conditional decisions needed to cover all bases. I'll get there though and am looking forward to this next stage in my trading journey.

I can now better appreciate those of you who have written EAs or other programs from scratch. Kudos to you. :)

Ps. Question for VBA programmers. In terms of processing usage and efficiency is it better to use more or less modules? As a complete noob, I'm finding it simple to write lots of little modules that do specific tasks and then write an overlay kind of module that calls the little ones as needed - but I'm not sure if that taxes the PC's processing too much?!

One tool that can greatly assist a programmer dealing with logic issues is the Karnaugh Map. You can use these to simplify expressions of multiple conditions into just a few lines of code. Learning to use these makes life sooo much easier when dealing with, say, long entry, long exit, short entry and short exit signal combinations.

As for VBA modules. IMHO if you are only ever going to use the functionality of that piece of code one just write the code inline. If you're going to use the functionality in a few different applications then write a module and save it so it can be reused. The difference on processor efficiency between inline and modular code is almost irrelevant; what makes the difference is the ease at which you can reuse and debug the code. If you write a module, you can write a small application that extensively tests the module so you know it works. When you call that function in another program and that program crashes, you can be pretty sure your well-tested module is not to blame and it must be another cause. (Most programmers don't write test apps for their modules; I do it religiously. The short time taken to write a proper test-rig for the function will save many hundreds or thousands of hours bug-hunting later!)

Hope this helps.


wabbit :D
 
One tool that can greatly assist a programmer dealing with logic issues is the Karnaugh Map. You can use these to simplify expressions of multiple conditions into just a few lines of code. Learning to use these makes life sooo much easier when dealing with, say, long entry, long exit, short entry and short exit signal combinations.

Okay. I had a quick google for Karnaugh Map and it's a little beyond me right now but will probably make more sense down the road.

As for VBA modules. IMHO if you are only ever going to use the functionality of that piece of code one just write the code inline. If you're going to use the functionality in a few different applications then write a module and save it so it can be reused.

That makes sense. I've kind of done the multiple modules by default as I started out by using the macro recorder for buy, sell, cancel, close etc for manual trading which I then mapped to buttons. And that's made life simple now as I can just call those modules from the auto trading program, which I'm writing from scratch.

If you write a module, you can write a small application that extensively tests the module so you know it works.

Can you expand on this a bit more please? What I've been doing is have a module do something arbitrary as a test, which I can then replace with the real decision once I know it's working.

For example: I might want to place a buy given X being true. So to test that module, if X then print the current time in a certain cell or write "Hello" somewhere. So I know that part of the overall tree is working. Then when I live test, I just replace that outcome with order placing code instead.

But I get the feeling you might mean something else. :confused:


The short time taken to write a proper test-rig for the function will save many hundreds or thousands of hours bug-hunting later!)

Given how painful I'm finding this process, I can really appreciate your enthusiasm for such an idea.

Thanks wabbit.
 
Can you expand on this a bit more please? What I've been doing is have a module do something arbitrary as a test, which I can then replace with the real decision once I know it's working.

For example: I might want to place a buy given X being true. So to test that module, if X then print the current time in a certain cell or write "Hello" somewhere. So I know that part of the overall tree is working. Then when I live test, I just replace that outcome with order placing code instead.

But I get the feeling you might mean something else. :confused:

Let's pretend you have a VB module that divides one input argument by another input argument. i.e.

private function Divide(single x, single y) as single
{
Divide = x/y
}

Your test application will call this function with maybe a few (thousand) values and test the outcomes. You would test with all sorts of combinations such as the largest single positive value, largest negative single value, 1, and few other numbers somewhere in the range; test with numbers larger than than single data type and see that errors are correctly handled, test for divide by zero errors again to see that errors are correctly handled.

Each time you get an error in your module, you need to deal with it. Eventually, after a lot of testing your module will be, for all intensive purposes, bullet-proof.

Along the way, you will learn to identify different errors which can occur when dealing with different data types, and human operators! And how to write good testing applications. Once you have a decent test-rig, you can apply it quite widely, meaning you don't necessarily have to write a whole new testing routine for each and every module (although, this is what they teach at Uni!)


If you want more help with Karnaugh mapping logic and decision trees, just give me a shout.


Hope this helps.

wabbit :D
 
initial testing of version 2 of the derwent (named after the derwent river - which wasnt dammed and was allowed to 'go with the flow' ... ie the trend).

this is for just a single currency. im expecting this system to work across both multiple currencies and multiple markets, as it looks for a strong trend and buys into it at an opportune time.

time frame = 2004-2008.

no compounding returns/money management
 

Attachments

  • TesterGraph.jpg
    TesterGraph.jpg
    67.1 KB · Views: 7
Thanks wabbit. No doubt that was simple to you, but I'll need to read it a few times and try it out to make sense of it. :) But thanks!
 
Thanks wabbit. No doubt that was simple to you, but I'll need to read it a few times and try it out to make sense of it. :) But thanks!

if u dont test youll end up with a divide zero error or something which will wipe out your trading account in 5 minutes.

not wanting to be alarmist.
 
Top