Phoenix 2007 (new thread) | Page 45 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phoenix Forex Manager trading journal 25 replies

Phoenix Foundation 1 reply

Can someone fix the Phoenix EA? 0 replies

Automated Trading Championship 2007: New Challenge! 33 replies

Phoenix is here! 848 replies

Attachments: Phoenix 2007 (new thread)
Exit Attachments

Phoenix 2007 (new thread)

  • #881
  • Nov 16, 2006 11:47am Nov 16, 2006 11:47am
Have just one question, more of IT related nature. What programme is that in your zip doc with .mq(?) extension. Can not open it at all. Mine is Explorer 7, is it coz of that?
  • #882
  • Nov 16, 2006 12:08pm Nov 16, 2006 12:08pm
Privet, Artemos. Can you please post your backtest with the new parameters for USD/JPY you offered. I've tested them, but the results are about 50% worse than with initial settings
  • #883
  • Nov 16, 2006 1:28pm Nov 16, 2006 1:28pm
Quoting abror_aa
Have just one question, more of IT related nature. What programme is that in your zip doc with .mq(?) extension. Can not open it at all. Mine is Explorer 7, is it coz of that?
I am not sure what zip document you are asking about. *.mq4 files in this forum are for MetaTrader which is a free charting and trading program. In this thread the file will probably be an expert advisor. You can use notepad to see the code in a *.mq4 file if you do not have MetaTrader.
  • #884
  • Nov 16, 2006 3:17pm Nov 16, 2006 3:17pm
hi Dagnar,

I just had a look at your idea...backtested it and - WOHOOO! - AMAZING!
Maybe I am doing something wrong but it's really good. I'll have a closer look @ it this weekend, I think. Looks great.
Thank you!

Quoting Dagnar
Great job, cool EA, my respect, Hendrick.

Trying some backtests of different phoenix versions, results differs from doubling (and more) deposit to margin call. Looking at indicator for this EA (dot't remember by who, posted at championship forum), I've got an idea - open trade only by some consecutive signals.

added some code to EA( like this for sell signal) :

if (CurTime()-s_time>900) {ss=ss+1; bs=0;}
if (ss>=signal_count) {
res=OrderSend(...SELL....);
ss=0; s_time=CurTime();}

I think, it will help to filter out single signals at gaps, etc..

Made this for ph_4 (contest ver) and for ph_5 at mode1 (last ver) - the result is a new significant parameter for optimizing using pref_settings=true; For EURJPY, for example, I've got 50% increase in profit turning signal_count to 5 (not less than five consecutive signals opens trade). Surprisingly, this didn't reduces total amount of trades, only placing orders more precise.

Attaching both changed files, hope it's usefull contribution to phoenix development
  • #885
  • Nov 16, 2006 5:10pm Nov 16, 2006 5:10pm
Quoting Dagnar

added some code to EA( like this for sell signal) :
Dagnar, I've posted your msg and your changed version here, at sisters thread:

http://www.forex-tsd.com/expert-advi...html#post64083

in case if you are not aware of it.

I hope I didn't violate some rules.
  • #886
  • Nov 17, 2006 1:52am Nov 17, 2006 1:52am
Mind this:
!!!!!
if(lot>10) lot=10;
!!!!!

Because, in real trading there can be problems to open a trade with a lot more, than 1'000'000= (1/100)! It can be detained and open at over price or cancelled at all!
Attached File(s)
File Type: zip Test.zip   15 KB | 452 downloads
  • #887
  • Nov 17, 2006 2:28am Nov 17, 2006 2:28am
Quoting tradeigel
hi Dagnar,

I just had a look at your idea...backtested it and - WOHOOO! - AMAZING!
Maybe I am doing something wrong but it's really good. I'll have a closer look @ it this weekend, I think. Looks great.
Thank you!
Can you post the statemant?
  • #888
  • Nov 17, 2006 5:10am Nov 17, 2006 5:10am
Appeal. Dear Hendrick be kind say the opinion concerning the offered introductions to the expert. And how we all must correct Phoenix 5 Mode 3?
  • #889
  • Nov 17, 2006 9:01am Nov 17, 2006 9:01am
to your question about posting the statement:
No, I'm not gonna post the statement, back test it yourself. I am not saying that to offend you or sth. just but because your OWN results are way more valuable to you.

And to this "Appeal":
Quoting Oligarh
Appeal. Dear Hendrick be kind say the opinion concerning the offered introductions to the expert. And how we all must correct Phoenix 5 Mode 3?
I'm not sure what you're talking about but Hendrick left the forum so he probably won't answer you.
And btw, read the whole thread, please. That is one reason why Hendrick left this thread, I think, so many people asking questions which they wouldn't ask if they had read the whole thread. That is: unnecessary questions.
  • #890
  • Nov 17, 2006 10:08am Nov 17, 2006 10:08am
Quoting tradeigel
to your question about posting the statement:
No, I'm not gonna post the statement, back test it yourself. I am not saying that to offend you or sth. just but because your OWN results are way more valuable to you.

And to this "Appeal":


I'm not sure what you're talking about but Hendrick left the forum so he probably won't answer you.
And btw, read the whole thread, please. That is one reason why Hendrick left this thread, I think, so many people asking questions which they wouldn't ask if they had read the whole thread. That is: unnecessary questions.
Couldn’t agree more...<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o></o>
  • #891
  • Edited 11:10am Nov 17, 2006 10:59am | Edited 11:10am
great, Vince.

@ wackena and all the others:

If there is really such a big difference between demo and live we should try to compensate that. Is it because the price / quotes are different and therefore less high-quality signals are generated? If so, we could let one demo run and pipe its signals as 'Buy at market' for the live account. So if a demoSignal is generated then open an order live. But if the discrepancy is because of the requotes, slippage and spread widening, then it sucks. Then we can just choose the best broker and trade there.
Do you think there's a big difference between the contest's results which are demo and your live ones? If so, why is that? Is FXDD (which might provide the feed) better for Phoenix? In my backtesting FXDD shows better results than Alpari's data which I normally use for BT. With FXDD I get 90% m.q., too.
What do you think about these issues?
  • #892
  • Nov 17, 2006 12:23pm Nov 17, 2006 12:23pm
Quoting tradeigel
great, Vince.

@ wackena and all the others:

If there is really such a big difference between demo and live we should try to compensate that. Is it because the price / quotes are different and therefore less high-quality signals are generated? If so, we could let one demo run and pipe its signals as 'Buy at market' for the live account. So if a demoSignal is generated then open an order live. But if the discrepancy is because of the requotes, slippage and spread widening, then it sucks. Then we can just choose the best broker and trade there.
What do you think?
tradeigel,

As I feel most do, I use the present Strategy Tester and available data to help me develop EAs. Backtest bar data is basic OHLC data. LIve data would contain the interlace movements within each bar. Until LIve tick data is recorded, purchased or made available from Brokers, my confidence in Backtesting is moderate at best. As posted earlier, IBFX Broker plans to offer Live tick data in near fulture.

As important as I feel about using Live tick data to backtest, it is not my only concern. I also see difference from my Demo to Live results on several EAs, including Phoenix and Firebird. I've seen and have been told that after Demo charts are checked to confirm good entry points, that in Live trading more bad entry points are seen. I feel one reason is that Demo order execution is instant and Live trading can have delays many times between order request and order execution. One delay can be whether your Broker is providing tick data feeds direct or through a Trading Desk, and how they filter the data. Another reason is on low volume pairs, Live trades are not executed until minimum volume is accumulated. I don't know what is the minimum volume, but I have manually executed orders and they did not go though and FXDD advise this was the reason.

The slippage variable has no effect in Demo testing, but does on Live trading. Live trading can have either, if slippage too high, order execution to be many pips away from order request. Or miss the order completely, if slippage too low. You do not see these effects in Demo or backtest.

To try to summarize, I think using Live tick data for backtesting is critical, and at present, we will not see identical results between Demo and Live trading.

Everyone should ask their Broker if they have Live tick data feeds now for Strategy Tester. If not, do they plan to offer one soon. If someone finds a Broker providing Live tick data for MT4 Platform Strategy Tester, please post the Broker name.

Well, enough rambling on my part. Everyone have a great weekend.

Wackena
  • #893
  • Nov 17, 2006 12:31pm Nov 17, 2006 12:31pm
Quoting wackena
...Well, enough rambling on my part. Everyone have a great weekend.

Wackena
LOL. I must admit Wackena that your "Rambles" have been very inspirational. Keep on a ramblin buddy !!

Cheers,
Thom
  • #894
  • Nov 17, 2006 1:13pm Nov 17, 2006 1:13pm
I have been trying to figure out some settings or mode that is working well right now. I am not a programmer so I don't understand why Phoenix enters a trade when it does. When I manual trade I use a longer timeframe to give me a more accurate picture of what is happening. For some reason Phoenix has not performed well in the past week or so and we know it can't always win. Big problem is, it lost all the previous gains. My question is, can we use a longer timeframe as a guideline. I am sure there is a solution. It just seems a shame to me not to figure it out. Also if anyone has been doing well over the last week or so, please share your results and settings.

Thanks
  • #895
  • Nov 17, 2006 2:03pm Nov 17, 2006 2:03pm
Greetings from the Sister Thread on Forex-TSD for Phoenix. We have spawned off two threads to help manage the questions. They are titled:
"Phoenix - FAQ, Stable, User support -Read Post #1"
Phoenix - Development+Suggestions - MQ4 in Post#1
The first post of each thread will contain the most updated information - so the entire thread doesn't need to be read.

I have suggested and we are moving to a development style that works for very large open source projects. There are three basic kinds of users: people who want working code, people who want code to test with, and developers who want to include their ideas or fix existing bugs.
The first group will have a "stable" release (which we don't currently have a candidate for) that has no major bugs, works well and has gone through a great deal of testing.
The second group will be playing with different releases, experimenting with settings and enabling/disabling the new features.
The third group will be a compilation of most or all code submitted each week into a single file. If the group that includes the code doesn't include the code other people want, then there might be a code fork in development. I'm posting here to avoid having 30 different version of phoenix all with their own strengths and weaknesses combined with random naming conventions and no settings files.

Phoenix has very fragile settings, we need to use good settings in order to avoid absolute death of the EA in ?? weeks. I am considering a website that only manages settings files and test results organized by the version of Phoenix, etc.

I will attempt to merge the mode 3 fix that Hendrick endorsed and some of the other changes that Hendrick planned on for release later tonight.

How does this community feel about setting up sister threads for user and development (or keep this thread and make another dev-only thread)? For active developers, we also have a private mailing list so we can converse about code changes in private.

I am going to number the version released 5.6.4 as a weekly build, and try to make it a candidate for "stable". Once Mode 3 settings have been verified, it will move up the chain to the first "stable" release I hope.
  • #896
  • Nov 17, 2006 2:48pm Nov 17, 2006 2:48pm
@daraknor:

AWESOME, just awesome, this idea. Exactly what I was thinking of. I personally use linux, so I know this concept of an os-project, it has LOTS of potential!
Additionally, I would be up for the some of each part (1,2 and 3).
  • #897
  • Nov 17, 2006 5:13pm Nov 17, 2006 5:13pm
Quoting Dagnar
Great job, cool EA, my respect, Hendrick.

Trying some backtests of different phoenix versions, results differs from doubling (and more) deposit to margin call. Looking at indicator for this EA (dot't remember by who, posted at championship forum), I've got an idea - open trade only by some consecutive signals.

added some code to EA( like this for sell signal) :

if (CurTime()-s_time>900) {ss=ss+1; bs=0;}
if (ss>=signal_count) {
res=OrderSend(...SELL....);
ss=0; s_time=CurTime();}

I think, it will help to filter out single signals at gaps, etc..

Made this for ph_4 (contest ver) and for ph_5 at mode1 (last ver) - the result is a new significant parameter for optimizing using pref_settings=true; For EURJPY, for example, I've got 50% increase in profit turning signal_count to 5 (not less than five consecutive signals opens trade). Surprisingly, this didn't reduces total amount of trades, only placing orders more precise.

Attaching both changed files, hope it's usefull contribution to phoenix development

Great job Dagnar,

I believe this enhancement should be made into an updated version called Phoenix 6, since it works so well!
I've tested it and it works like a charm.
Does anyone else agrees with me?
Now we just need to figure out the best overall settings for eur/usd or if someone already has them, can you please share them.
  • #898
  • Nov 17, 2006 5:51pm Nov 17, 2006 5:51pm
Quoting wackena
As I feel most do, I use the present Strategy Tester and available data to help me develop EAs. Backtest bar data is basic OHLC data. LIve data would contain the interlace movements within each bar. Until LIve tick data is recorded, purchased or made available from Brokers, my confidence in Backtesting is moderate at best. <?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o></o>
Quoting wackena


Same here. Even though I do A LOT of back testing and optimizing I'm using the freely available ticks from alpari which are not the same as those used by my live trader. Therefore I'm not convinced if I'm wasting my time at the moment or if it has some use..
<o></o>

Quoting wackena
<o></o>
Quoting wackena
Everyone should ask their Broker if they have Live tick data feeds now for Strategy Tester. If not, do they plan to offer one soon. If someone finds a Broker providing Live tick data for MT4 Platform Strategy Tester, please post the Broker name.<o></o>
<o></o>

Had a little chat with my live broker this morning and hopefully they will provide this soon... How soon? No idea...<o></o>

<o></o>
At the moment I'm comparing my live ticks and quotes with the demo ticks provided with my demo accounts (which I used for all my tests).

So far so good BUT this doesn't guarantee at all the other issues razed by you (time it takes to execute a trade and so forth...).
But I think, however annoying (GRRR!!!) it might be to get 1 or 2 pip difference between live and demo execution, it might not be the nail to the coffin of Phoenix (ever the optimist..), since it's trades last quite a long time and over a big range...<o></o>

Good weekend!

Vince

(Will be abroad the next week so I will not be pestering the forum for a while ;-) )
  • #899
  • Nov 17, 2006 7:25pm Nov 17, 2006 7:25pm
Quoting ericbach
Great job Dagnar,

I believe this enhancement should be made into an updated version called Phoenix 6, since it works so well!
I've tested it and it works like a charm.
Does anyone else agrees with me?
Now we just need to figure out the best overall settings for eur/usd or if someone already has them, can you please share them.
I don't agree you should call it Phoenix 6. If you want to fork development and release your own copies as a separate code body, please choose a name variant instead of using the same name. How many Phoenix 6 can we make! I think it would be incredible if we could make a ton of partially working copies and call them all the same thing.
This is probably too sarcastic, but this post did come 2 posts after my plea to avoid doing this *exact* thing.
  • #900
  • Nov 18, 2006 1:35am Nov 18, 2006 1:35am
5.6.4 relesed with 1 major code fix, 1 enhancement to trade logic and a setting to be used in research have been added.
http://www.forex-tsd.com/phoenix/463...-post-1-a.html

All modes may be working now, please backtest all 3 modes and post results:
http://www.forex-tsd.com/phoenix/464...imization.html

Your weekly Live Testing results are needed!
http://www.forex-tsd.com/phoenix/465...g-reports.html

Thank you all for staying with us. Hendrick represented a great example to our forums, programming and trading. I hope the community that formed around a selfless person will endure, and Phoenix will improve as we grow and work together.
0 traders viewing now