002097's PwR scalping auto trading robot by Steve Hopwood | Page 18 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trying scalping with a 1000 bucks - Scalping is expensive! 39 replies

1M & 5M Scalping System (Scalping Paradise) 21 replies

Random Auto Scalping to Profits 36 replies

PwR EVA 16 replies

MoForce's PwR System 25 replies

Attachments: 002097's PwR scalping auto trading robot by Steve Hopwood
Exit Attachments

002097's PwR scalping auto trading robot by Steve Hopwood

  • #341
  • Sep 10, 2011 5:41am Sep 10, 2011 5:41am
Quoting ianp
Excellent 4 days, well done guys,

I am using the first version posted ootb,

Trading times 7-17 uk now time.

my forward test results are
Wins = 22
Loss = 9

If only this would happen very week !!!!!!!
Let's hope it does! I've been trading this system for almost a year with all the limitations a human being has, and yet has been profitable.
The development of this Bot is going to give me the chance to enjoy life again.
I don't know about you guys, but I find trading a pretty boring and lonely thing, that's why I'm falling in love for Aurora. She's hot!
  • #342
  • Sep 10, 2011 5:49am Sep 10, 2011 5:49am
Quoting SteveHopwood
Ok guys, moving on from some of the dafter posts I have just read:

Many thanks for the invaluable help that so many of you have given me this week. Without you, the bot would not be where it is now.

It looks as though we (and I really do mean 'we') have got the stealth stuff licked; I arrived home tonight to a bot that performed magnificently in my absence.

Having got the basics right, or at least appearing to - and who cares if they are not so long as the pips roll in - next thingy's on the horizon are:
[list][*]minimum trigger candle size....
I can't express how glad I am with the development of Aurora, I just want all of you to enjoy the profits of this neat system.
She's becoming a beautiful bot.

  • #343
  • Sep 10, 2011 5:50am Sep 10, 2011 5:50am
Three things I wouldn't normally do:-

 

  1. Open an account with FXDD (exorbitant spreads);
  2. Begin trading on a Friday afternoon;
  3. Let a robot run my trades: hell, I don't normally use indicators apart from fibs or MMLs.


6 wins, 1 loss, +2%.

I know it's not statistically significant in any way but it's a damn good start. I'm subscribed.

  • #344
  • Edited 5:59am Sep 10, 2011 5:50am | Edited 5:59am
Instead of using a fix candlesize, it could be worth a try to experiment with some priceaction (look for candlehigh / low, etc.).

In my manual trading I am experimenting with some indicators I wrote to see how I can improve my trading. That indicators give alarms whenever a valid signal occurs.

I post the priceaction-thing here, maybe it could be useful. It is a simple function which returns a string ("FLAT" / "LONG" / "SHORT").

Inserted Code
 
//+------------------------------------------------------------------+
//| function getPriceAction                                          |
//+------------------------------------------------------------------+   
// returns "FLAT"/"LONG"/"SHORT"
// call, e.g.: string priceAction = getPriceAction(symbol,timeFrame,i);
string getPriceAction(string symbol,int timeFrame,int shift)
{
 RefreshRates();
 double price = MarketInfo(symbol,MODE_BID);
 double openPrice = iOpen(symbol,timeFrame,shift);
 double highPrice = iHigh(symbol,timeFrame,shift);
 double lowPrice = iLow(symbol,timeFrame,shift);
 double openPriceB4 = iOpen(symbol,timeFrame,shift+2);
 double highPriceB4 = iHigh(symbol,timeFrame,shift+2);
 double lowPriceB4 = iLow(symbol,timeFrame,shift+2);
 double closePriceB4 = iClose(symbol,timeFrame,shift+2);
 string priceAction = "FLAT";
 if (price > openPrice
     && price > openPriceB4
     && price > closePriceB4
     && lowPrice > lowPriceB4
     && highPrice > highPriceB4
     //&& lowPrice > openPriceB4 test
     ) priceAction = "LONG";
  else if (price < openPrice
           && price < openPriceB4
           && price < closePriceB4
           && lowPrice < lowPriceB4
           && highPrice < highPriceB4
           //&& lowPrice < openPriceB4 test
           ) priceAction = "SHORT";
 
 return(priceAction);          
}

edit: added RefreshRates(); (in my indi, I call that before the functioncall).
PM me with coding requests and I'll probably put you on my ignore list
  • #345
  • Sep 10, 2011 5:53am Sep 10, 2011 5:53am
Quoting SteveHopwood
There is a lot of twittering about CCI here.

Just so you guys know. I am going to do nothing about it.

Post a few weeks successful live manual trading results that include CCI in the mix and prove the difference that trading with it and without it made, and I will take some notice. Until then, stay stum, OK?

I am not going to allow this thread to descend into the anarchy of, "How about this indi."

There's no need to repair something that isn't broken.
  • #346
  • Sep 10, 2011 5:56am Sep 10, 2011 5:56am
Quoting dragosd1
Nice as always...thanks!

Btw, don't get yourself too happy, I need some pieces of code for myself too, hahahaha! But can wait a bit.

Regarding Aurora, it's fine, I can't wait for that new release. Will test the hell out of it! Meantime I'm gonna test other pairs, then, with the latest version.

Have a good night!

P.S. Hmm, this baby needs volatility, so I'm gonna check also the hours when the wins-loss have produced, will be interesting.
I'm going to work on that too as soon as I can (currently on vacations...) Back to real life on Monday.
  • #347
  • Sep 10, 2011 6:09am Sep 10, 2011 6:09am
Hi FxDayTrader

in effect this code is already in Aurora, it is conceptually the same as the check that the trigger candle has a HH & HL for long than the previous candle, and that the trigger candle has a LL & LH for short than the previous candle

the main difference though is your code does not check the previous candle, it checks the candle before that, ie. compares candles 1 & 3 (and ignores 2)

and also checks the current price against those values (which for us would not be a good thing as we wish to enter as as soon as we pass the low or high of the trigger candle).

Regards, John

Quoting fxdaytrader_
Instead of using a fix candlesize, it could be worth a try to experiment with some priceaction (look for candlehigh / low, etc.).

In my manual trading I am experimenting with some indicators I wrote to see how I can improve my trading. That indicators give alarms whenever a valid signal occurs.

I post the priceaction-thing here, maybe it could be useful. It is a simple function which returns a string ("FLAT" / "LONG" / "SHORT").
  • #348
  • Sep 10, 2011 6:58am Sep 10, 2011 6:58am
anybody running this ea for 24x5
  • #349
  • Sep 10, 2011 9:52am Sep 10, 2011 9:52am
It is play time gals and galesses.

Latest update in post 1:

 

  1. a bit of housekeeping - moving some of the code to different locations to make stuff like the round number filter work.
  2. added the minimum candle size filter.
  3. added mptm-style trade management. Read the stealth section in post one for details.

So, would some of you coders care to run a few visual backtests to see if you can spot anything wrong.

A question for those here with more logical minds than mine - most of you, in other words. Most of the functions (closure and stop loss manipulation) work off the Bid price as this is the one that displays on the charts by default. Buy trades are triggered by the Ask, so on a chart it looks as though they have opened prematurely.

Do you guys think that all long-trade decisions should be made off the Ask, or should all trade decisions be made off the Bid?

Have fun.

  • #350
  • Sep 10, 2011 9:55am Sep 10, 2011 9:55am
Quoting laharjaya
anybody running this ea for 24x5
I have been running it on my coding demo 24/5 and it has done ok. I have lost track of my running total, but the account balance is rising at a satisfying speed.

Running it live 24 hours a day would probably not be a good idea. The crims understand that traders need to sleep, so 24/5 trading would be a dead giveaway that we are using an EA.

  • #351
  • Edited 10:07am Sep 10, 2011 9:57am | Edited 10:07am
Steve

Although I have posted as quoted below in response to a post which may have concerned the issue of correct downloads this was not the reason for my post nor for this response, it was only a trigger. Therefor I am not engaging in a discussion of downloads or their failure.

Quoting NicoScholtz
A suggestion is for Steve to include an incremental revision number for each release or version in the copyright property as it is easily readable in both mq4 and ex4 format with a normal text editor such as MS Notepad. If this is applied with discipline it would make it unambiguous as to which version is in use, which would help with debugging and support.
Inserted Code
 
#property copyright "Copyright © 2009, Steve Hopwood, V1.0XX"
I suggested a fairly simple method, there may be other easier and/or better ways, to help improve Aurora by assisting in debugging and monitoring of performance by referring to an identifiable version. This is the only reason for the original post and for this response.

As illustration the quoted section below refers to a short selection of comments which I have extracted from posts on this thread. These are not in any particular sequence.

Quote
"I have used last version, in post #1..."
"For some reason, the new version..."
"I have added back to the new Aurora..."
"Also, I have the pre-stealth Aurora version on..."
"I am using the first version posted ootb.."
"i ran aurora with ootb settings.."
"I got the version running just before the hidden stuff was added..."
To my mind it is very difficult, if not impossible, to relate the comments or results to a specific version because there is no way to indentify or link it. This makes performance comparison and debugging more difficult.

Imagine the above snippets were as follows:
Quote
"I have V1.005..."
"For some reason, V1.003..."
"I have added back to Aurora V1.005..."
"Also, I have the V1.004 Aurora on..."
"I am using version V1.007 ootb.."
"i ran aurora V1.004 with ootb settings.."
"I got V1.005 running..."

I am asking for a decision about this based on logic and merit only.
I am not spoiling for a fight and hereby rest my case.

Regards, Nico
Success is a marathon, not a sprint!
  • #352
  • Sep 10, 2011 10:13am Sep 10, 2011 10:13am
My 2c (or as you are in the uk my 2p) on the bid / ask question

this is always a nightmare, people usually look at designing systems by looking at a chart and going if x + y happens I enter at z

therefore by design as charts are based on bid the system is also naturally at bid

the problem if you use Ask as the trigger (you of course open the order at ask) is that when someone is visually looking at the chart and going, hmm, the candle didn't breach the previous candles high - why did it enter, well quite simply it is because the Ask did (and as we have no tick based charts in MT4 it is not easy to see when there was spread spikes)

From my side, I find it easier when using pending or pending type (ie. hidden) orders it is always easier to use the values that the user can easily verify on the chart at a later point in time (ie. Bid),

of course when you look at the chart when the long trade is open, it will look as though it entered late (because it entered at ask), but at least when back verifying trades made, it is easier to see why trades entered, ie. you can see that the high of the entry candle was higher than the high of the trigger candle.

hmm, not sure of the exchange rate, maybe this was a little more than 2p worth - perhaps 3p

Regards, John

Quoting SteveHopwood
It is play time gals and galesses.
A question for those here with more logical minds than mine - most of you, in other words. Most of the functions (closure and stop loss manipulation) work off the Bid price as this is the one that displays on the charts by default. Buy trades are triggered by the Ask, so on a chart it looks as though they have opened prematurely.

Do you guys think that all long-trade decisions should be made off the Ask, or should all trade decisions be made off the Bid?

Have fun.

  • #353
  • Sep 10, 2011 10:23am Sep 10, 2011 10:23am
Steve

Thanks a lot for the new version. Will test of course!
Quoting SteveHopwood
Do you guys think that all long-trade decisions should be made off the Ask, or should all trade decisions be made off the Bid?
Since you asked, my vote would be that decisions for longs are taken off Ask as this is how it happens in real life. People reading this thread should know the difference.

Regards, Nico
Success is a marathon, not a sprint!
  • #354
  • Sep 10, 2011 10:31am Sep 10, 2011 10:31am
John

Your post makes good sense for just keeping everything at Bid, especially when looking at history. Maybe I should just delete my previous post...

So, my final vote goes for...........

Let the author decide what is easiest to code and maintain!

Quoting jmw1970
My 2c (or as you are in the uk my 2p) on the bid / ask question
Regards, Nico
Success is a marathon, not a sprint!
  • #355
  • Sep 10, 2011 10:49am Sep 10, 2011 10:49am
OOTB settings:

4 loss
9 win

I'll continue with demo for couple weeks...then going live!
Attached Image (click to enlarge)
Click to Enlarge Name: fx1.gif Size: 33 KB
  • #356
  • Sep 10, 2011 12:32pm Sep 10, 2011 12:32pm
Hi,

I'm also following the manual trade thread (http://www.forexfactory.com/showthre...292877&page=17) and I have compared your chart from yesterday (9/9 2011 - grats on a good day btw) with the current version (not the one just posted above) of Aurora ootb with the only change being: hiddenpips = 0.

Here is the result (as well as attached)

Manual:
8 wins, 3 losses.
15% - whatever that may be.

Aurora:
5 wins, 3 losses.
Very tiny plus of €50.54 with 1 lot.

Discrepancies:
Trade 1: Aurora nothing, manual lose.
Trade 4: Aurora nothing, manual lose.
Trade 6: Aurora lose, manual win.
Trade 9: Aurora nothing, manual lose.
Trade 11: Aurora lose, manual win.

Trade 1 may be due to difference in high/low between brokers?

Trade 4 was not triggered by Aurora for some reason even though it looks like the parameters was met? PwR is going lower and LH/LL is also true.

Trade 6 is triggered too early in the manual trade as PwR is not going lower between the topmost bull-candle and the first bear-candle.

Trade 9 is perhaps due to spread? I normally only have between 0.4 and 0.8 pip though.

Trade 11 is triggered too early in the manual trade per your own rule of disregarding candles > 15 pip.

I hope you can elaborate a little on the charts so perhaps Aurora can be better? Or I can

And no, its not profitable having it run 24/5. I did that from Tuesday morning to Thursday evening and my demo went down more than 70%

Have a nice weekend!
Attached Image(s) (click to enlarge)
Click to Enlarge Name: autotradebyaurora.jpg Size: 193 KB
Click to Enlarge Name: manualtradedby002097.gif Size: 21 KB
Click to Enlarge Name: resultstradedbyaurora.jpg Size: 82 KB
  • #357
  • Sep 10, 2011 12:44pm Sep 10, 2011 12:44pm
I tested aurora yesterday 9th Sep between 8 - 18 CET on alpari. All default settings
7 wins 1 loser
  • #358
  • Edited 1:33pm Sep 10, 2011 1:03pm | Edited 1:33pm
Quoting lizu
I tested aurora yesterday 9th Sep between 8 - 18 CET on alpari. All default settings
7 wins 1 loser
Weird how much difference there is between the results. Could you perhaps post a chart marking your trades?

Edit: I may have found why mine is off by so much. I'm sitting here Saturday evening (I have work tomorrow at 5am ) where the spread is 2 pip and running the Friday through the strategy-tester in MT4.
I'm to tired to figure out if that accounts for all of the differences.

Good night!
  • #359
  • Sep 10, 2011 6:30pm Sep 10, 2011 6:30pm
Quoting 0002097
There's no need to repair something that isn't broken.

I have to agree , if it works , just let it make pips !!!
  • #360
  • Sep 10, 2011 10:00pm Sep 10, 2011 10:00pm
Awesome work once again Steve...

Can't wait for next week to let this baby rip.....
0 traders viewing now