stevehopwoodforex.com • The Great HGB Foward Test

Author:

trendinator [ Wed Jan 28, 2015 9:03 am ]

Post subject:

The Great HGB Foward Test

WhoKnows » Wed Jan 28, 2015 5:58 pm wrote:
Ouch..... ... that would be bad.
I thought it would go to minutes as base... I guess I'll better try what happens or the 'coder/tech specialists' might know the definitive answer..

Maybe rephrase it as third and fourth question:
- Would it not make sense to make HGB work on one minute bases?
- And would a minute base have an effect on very temporary fraction of a second signals that disappear?

it already works on 'minute bases', as in, it checks HGI for a signal every 60 seconds by default, by HgiReadDelaySeconds. the effect of this has been discussed here:

http://www.stevehopwoodforex.com/phpBB3 ... 80#p111980

edit: fixed link. here's what i said:

trendinator » Tue Jan 20, 2015 11:44 pm wrote:im curious about the following line:
Code: Select all
extern int      HgiReadDelaySeconds=60;//cpu saver. Read the indi every minute by default
lets look at the following scenario:

12:00:00 - HGB started - reads HGI, no signal

12:00:30 - there is a signal, but it quickly goes away after a few seconds

12:01:00 - HGB reads HGI again - no signal = there are no trades taken here.

Let's say I had started HGB at 12:00:30 it would have seen the signal and taken the trade?

so is it possible that the time of setup can have an effect on the trades taken?

stevehopwoodforex.com https://www.stevehopwoodforex.com/phpBB3/ Print view

The Great HGB Foward Test https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4044	Page 20 of 27

Author:	barry [ Wed Jan 28, 2015 4:34 am ]
Post subject:	The Great HGB Foward Test
ok sounds like a good idea. perhaps for the next test we could create a profile and get everyone to load that? would that work? Yes, good one, that's a much better idea

Author:	Gertje [ Wed Jan 28, 2015 6:56 am ]
Post subject:	The Great HGB Foward Test
WhoKnows » Wed Jan 28, 2015 12:41 am wrote:Just two theoretical questions: - Would it help to use a TradingTime setting, even if it is like -10.00,+10.00 to synchronize instances. I ask this because I used TT settings in my tests (described in the BoHGB thread) and saw that different instances often executed trades almost on the same millisecond. - As this is certainly no scalping EA, and we have ReadDelay of 60 seconds, why do we run this EA on ticks? What would happen if you run it on 'EveryTickMode=false'? Is there any reason that we need the Tick mode? I might be miles off target, but in my understanding 'EveryTickMode=false' makes the EA only trade at candle close. On a H4 TF this is..... Err.... Less desirable.

Author:	Gertje [ Wed Jan 28, 2015 6:58 am ]
Post subject:	The Great HGB Foward Test
rapple » Wed Jan 28, 2015 1:03 am wrote: Should we also have everyone on the same version of Empty4. Steve does not guarantee his code on the later builds especially 7xx. I am running Version 4.0 Build 645. Steve did post the terminal.exe for the version he uses but I am not sure where And then when you go live, find out MetaQuotes has disabled trading on older versions.... I always learned you're supposed to practise the way you want to go to battle.

Author:	WhoKnows [ Wed Jan 28, 2015 7:58 am ]
Post subject:	The Great HGB Foward Test
Gertje » Wed Jan 28, 2015 4:56 pm wrote: WhoKnows » Wed Jan 28, 2015 12:41 am wrote:Just two theoretical questions: - Would it help to use a TradingTime setting, even if it is like -10.00,+10.00 to synchronize instances. I ask this because I used TT settings in my tests (described in the BoHGB thread) and saw that different instances often executed trades almost on the same millisecond. - As this is certainly no scalping EA, and we have ReadDelay of 60 seconds, why do we run this EA on ticks? What would happen if you run it on 'EveryTickMode=false'? Is there any reason that we need the Tick mode? I might be miles off target, but in my understanding 'EveryTickMode=false' makes the EA only trade at candle close. On a H4 TF this is..... Err.... Less desirable. Ouch..... ... that would be bad. I thought it would go to minutes as base... I guess I'll better try what happens or the 'coder/tech specialists' might know the definitive answer.. Maybe rephrase it as third and fourth question: - Would it not make sense to make HGB work on one minute bases? - And would a minute base have an effect on very temporary fraction of a second signals that disappear?

Author:	trendinator [ Wed Jan 28, 2015 1:44 pm ]
Post subject:	The Great HGB Foward Test
ok made some progress in my own tests today... seems like the number of bars in the history would've really had an effect on HGI and therefore HGB when setting up a new instance. i setup a new instance and tried this: http://www.stevehopwoodforex.com/phpBB3 ... f=15&t=254 unfortunately it doesn't always work and i found that I had anywhere from 100 to 2000+ bars in the H4 timeframe. it would actually report success when bars were more than zero, so 100 bars is a success, but its actually a total fail for HGI (we need 550). also it didnt compile properly in build 765 so ive tweaked the script for build 765 and specifically for HGI, so now it will check that there are at least 550 bars in the history on the H4 timeframe. here are a new set of instructions that should hopefully set you up with 28 charts running HGB and with more than 550 bars. if you like you can test this method for me. everything is in the zip. read the instructions! also... READ THE INSTRUCTIONS! HGB forward test V3 (new instructions).zip *see post one*

Author:	milanese [ Wed Jan 28, 2015 1:57 pm ]
Post subject:	The Great HGB Foward Test
trendinator » Wed Jan 28, 2015 1:44 pm wrote:ok made some progress in my own tests today... seems like the number of bars in the history would've really had an effect on HGI and therefore HGB when setting up a new instance. i setup a new instance and tried this: http://www.stevehopwoodforex.com/phpBB3 ... f=15&t=254 unfortunately it doesn't always work and i found that I had anywhere from 100 to 2000+ bars in the H4 timeframe. it would actually report success when bars were more than zero, so 100 bars is a success, but its actually a total fail for HGI (we need 550). also it didnt compile properly in build 765 so ive tweaked the script for build 765 and specifically for HGI, so now it will check that there are at least 550 bars in the history on the H4 timeframe. here are a new set of instructions that should hopefully set you up with 28 charts running HGB and with more than 550 bars. if you like you can test this method for me. everything is in the zip. read the instructions! also... READ THE INSTRUCTIONS! the HGI indicator itself has a history download function integrated, that will work as intended is the history is avaible on the broker server and you have not limited the max bars.. Cheers Tommaso

Author:	trendinator [ Wed Jan 28, 2015 2:10 pm ]
Post subject:	The Great HGB Foward Test
milanese » Wed Jan 28, 2015 11:57 pm wrote: the HGI indicator itself has a history download function integrated, that will work as intended is the history is avaible on the broker server and you have not limited the max bars.. Cheers Tommaso Well it didnt seem to always work when i tested it! I had plenty of charts with less than 550 bars. That was using hgi 15.0.6 and build 765

Author:	milanese [ Wed Jan 28, 2015 2:16 pm ]
Post subject:	The Great HGB Foward Test
trendinator » Wed Jan 28, 2015 2:10 pm wrote: milanese » Wed Jan 28, 2015 11:57 pm wrote: the HGI indicator itself has a history download function integrated, that will work as intended is the history is avaible on the broker server and you have not limited the max bars.. Cheers Tommaso Well it didnt seem to always work when i tested it! I had plenty of charts with less than 550 bars. That was using hgi 15.0.6 and build 765 yes it can fail, as it will only do 2 re-attempts, for not using too much time in the initialisation, and it will only load history of the TF and Symbol() the indicator is attached... Cheers Tommaso

Author:	Maximilian [ Wed Jan 28, 2015 6:23 pm ]
Post subject:	The Great HGB Foward Test
I got this shitload of trendwave trades again closing after 2seconds. Using HGB 2r ootb at the moment. When reinitialized HGB on the very same chart it stops. I had no errors in my logs and my VPS wasn´t maxed out either. I don´t know. Anybody else?

All times are UTC	Page 20 of 27
Powered by phpBB® Forum Software © phpBB Limited