"Phoenix - FAQ, Stable, User support -Read Post #1" - page 10

 
m6m6:
Should the default setting in 5.7.0 be change for 1.0 1.0 1.7 or should we continue forward testing with 0.8 1.0 1.2 ?

Either setting is fine, 1.0 1.0 1.7 was found in backtesting as the best but most of the Mode3 logic is disabled (jumping from 3 to 1 trade.) I did some quick testing and found 0.8 1.0 1.2 to be decent performance, while 0.5 1.0 1.5 quickly lost a lot of money (not enough benefit from wins that reverse at 0.6).

If someone wants to have the functionality of 5.6.8, Then I recommend 1.0 1.0 1.7 because it is the most similar setting to the way 5.6.8 worked. I don't recommend trading buggy software, (5.6.8) but that is me.

Test whatever you like!

I'm off to spend time with friends and family! I hope you do the same.

 

Merry Xmas!

daraknor:
Either setting is fine, 1.0 1.0 1.7 was found in backtesting as the best but most of the Mode3 logic is disabled (jumping from 3 to 1 trade.) I did some quick testing and found 0.8 1.0 1.2 to be decent performance, while 0.5 1.0 1.5 quickly lost a lot of money (not enough benefit from wins that reverse at 0.6).

If someone wants to have the functionality of 5.6.8, Then I recommend 1.0 1.0 1.7 because it is the most similar setting to the way 5.6.8 worked. I don't recommend trading buggy software, (5.6.8) but that is me.

Test whatever you like!

I'm off to spend time with friends and family! I hope you do the same.

Great explaination!

i will start to run live account with Phoenix after new year.

will post my results later.

Gambatte! (Good luck in japanese)

 

Version 5_6_06

Hi,

The version 5_6_6 modifies the order of missed form, could verify the code of the EA. The modification of the StopLoss confers order 12 after.

Thanks

Files:
 

Backtesting and live testing

I seem to be spinning my wheels here. I optimized on GPUSD with version 570, then ported the settings to 571. When I run the new settings on the same time period I get completely different results. Starting over with 571, but now I wonder whether I completely misunderstand the way the EA works in testing.

Hendrick says use PrefSetting false for (back)testing and manually set values of inputs. For live or demo forward testing, PrefSettings should be false, and the values from the code will be used. But then why bother doing backtesting if the values are going to be ignored when you do forward testing or go live?

Also, with MM=false, is it not possible to set a lot size other than 1?

Please clarify. We need a clear, well-written manual to explain the EA's innerds and how to optimize for best results. Confused!

 

We want the manual to be based on 5.7.1 because it is so much easier to understand. There are so many settings they needed to be organized a little.

I modified 5.7.1 a little to include a 'hidden bug' fix. I will 'undo' the fix and 'redo' the fix to make sure the trades are the same with the default settings.

The settings you get by default are fairly optimized, except for EURUSD. The EURUSD settings were included because they are better than the default (no currency specific), but they are still as good as I want.

PrefSetting means "do i use the values that are set up for specific currencies?" If true, it loads currency specific settings. If false, it loads the settings you specify.

 
Heder:
Hi,

The version 5_6_6 modifies the order of missed form, could verify the code of the EA. The modification of the StopLoss confers order 12 after.

Thanks

5.6.6 was known to be unstable. Please use the stable version 5.7.0 unless you want to help with testing.

 
daraknor:
We want the manual to be based on 5.7.1 because it is so much easier to understand. There are so many settings they needed to be organized a little.

I modified 5.7.1 a little to include a 'hidden bug' fix. I will 'undo' the fix and 'redo' the fix to make sure the trades are the same with the default settings.

The settings you get by default are fairly optimized, except for EURUSD. The EURUSD settings were included because they are better than the default (no currency specific), but they are still as good as I want.

PrefSetting means "do i use the values that are set up for specific currencies?" If true, it loads currency specific settings. If false, it loads the settings you specify.

A manual would be a great thing, even if it only involved descriptions of inputs initially. Perhaps a table of input settings should be done initially so that people know where to start with each currency. If you send me the raw data with annotations I could prepare a PDF. We can do that offlist if you wish.

PrefSettings: "do I use values that are set up for specific currencies?" Where does it get these settings?

I presume that the "settings you specify" are those we enter manually or prepare as a .set file. So if we optimize and create a .set file, we must always have PrefSettings=false (for optimization, backtesting, forward testing on a demo, and live trading). Is that correct?

What about the behaviour of MoneyManagement?

Thanks for your patience and assistance. Let me know what I can do to help.

BTW, I'm working on a potentially lossless optimization. Probably low yield, but maybe good for those of us who are sore losers.

 

I've done some reading on the subject.

Actually I've been reading both this & the dev threads, and all of the 2007 thread this weekend.

We need to organise a table with the following headder:

Ver name ; stat(stable/testing) ; pref currencies ; TF ; test results ; comments (how to operate it).

On that mater, could I open (drag&drop) phoenix on only one pair, some of them or should I use it on all of the pairs for it to be successfull? (the booster version had a very complicated set of instructions...).

BTW: I've read it earlier today, but I cant find it again. What is the main difference between mode 1, 2 & 3?

BTW 2: How does the MM function works in Phenix? does setting it to 0.003 means that you risk 3% of your equity in each deal? When does it stop? When your Core equity is down to 50%? 25? when?

 

Manual

daraknor:
We want the manual to be based on 5.7.1 because it is so much easier to understand. There are so many settings they needed to be organized a little.

Hi Daraknar,

Sorry for my absence but have been busy at work (yep, some people have to work during the holidays :-( )

Regarding the manual, I'm happy to adjust or even totally re-write my original version that was intended for Phoenix 4.

Having said that, I'm

1) not sure the original manual was any good in the first place

2) still waiting for a final stable version of Phoenix to be announced

I thought version 5.7.0 would be this but as usual, the development is continuing. I think it a bit pointless to write a manual and than having to make a completely new one a week later...

Suggestions?

Vince

 

Manual or auto?

I think that a manual is a good idea as long as it is understood as a set of rough guidelines. It could consist of two parts. One on general use of the EA, and another on the meaning of the inputs, perhaps in tabular form, as before. For example, I have no clue what the Grace inputs are about, or even whether to use them. We might also consider a section on optimization technique and backtesting, as I am sure that many may be confused about this (for example, me). E.g., does 99% backtesting on tick data represent a significant advantage over 90% backtesting on M15 timeframe data? While we're at it, how about a preamble about the general philosophy behind the EA, i.e., how its approach differs from others?

It does not have to be a finished product, since the EA isn't finished either. But some working guidelines would be most helpful.

Reason: