tonyharmony » Wed Feb 04, 2015 2:59 pm wrote:

boinvest » Tue Feb 03, 2015 8:14 pm wrote:

ALoveSvpreme » Tue Feb 03, 2015 8:50 am wrote:I keep getting this times infinity... any ideas?

Yes,
I was to much in a rush putting together previous version.
So attached is 2u.
Adittionally:
fixed the cashtakeprofit like in the HGB
fixed getnext20 like in the HGB
added lotsize check in order send function

G'day Boinvest, how are you?

Finally found you at last! Your own thread? Great news as I have the HGB v2o from the HGB forum, and was keen to know where you were going from there?

Is the above 2u in post 1?

Kindest regards

Tony

snailbeard » Sun Feb 01, 2015 5:01 pm wrote:Hi Boinvest,

Great work going on with the HGB EAs and I spent a couple of hours playing with the official version and your variant, to see if I could understand how they would perform over long periods of time. To keep it simple I turned off most trading methods except trend-trading.
I tried to keep the settings for both EAs as close as possible.

This is not a comparison between the two EAs I'm posting in case the developers find anything useful from the tests I did.

I have to get back to working on my own EA's so I'll just present a couple of screen shots.

Running the official HGB I found buy-stops when there was a big red arrow pointing down on the corresponding chart (GBPUSD H4). When I did a similar run for boHGI I found sell-stops a few bars after the big red arrow.

I was able to run HGB with 5 months of crappytester history but I was not able to run boHGB for longer because of different kinds of errors (A: volume mismatch; B: lot size error 131)

Official HGB:


BoHGB for the same period:


I tried to work around these errors by running on different brokers; I gave up with every-tick mode because of the first error and I tried different time-frames but decided to call it a day
and here is an extract from the log about volume issues:

Code: Select all

2015.01.31 20:18:17.141	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: GBPUSD BoHGB-2t Stop 0 price 1.676 take 1.673 error(131): 
2015.01.31 20:18:17.141	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: GBPUSD BoHGB-2t OP_SELLSTOP order send failed with error(131): invalid trade volume
2015.01.31 20:18:17.141	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: Alert: GBPUSD BoHGB-2t OP_SELLSTOP order send failed with error(131): invalid trade volume
2015.01.31 20:18:17.141	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: OrderSend error 131
2015.01.31 20:18:17.141	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: Alert: Screen snapshot taken 1 open ChartScreenShot.png
2015.01.31 20:18:17.073	2014.08.11 12:15  BoHGB-2t GBPUSD,M15: open #1 sell stop 0.10 GBPUSD at 1.67800 tp: 1.67300 ok
2015.01.31 20:18:17.049	2014.08.01 00:00  BoHGB-2t GBPUSD,M15: Checking GBPUSD for complete data.... number of M15 bars = 1001

[/size]

Anyway I hope this information is useful to the developers.

Wiggy1965 » Thu Feb 05, 2015 9:25 am wrote:
Hi Snail,

I thought the Holy Grail Indicator uses a form of the CSS (Currency Slope Strength) as part of it's logic....Based on this, I would have though back testing was impossible as the EA with HGI Indicator could not see the other currencies in order to determine strength....I might be wrong, but i have not considered doing back testing for that reason....

Wiggy

Wiggy1965 » Thu Feb 05, 2015 10:23 am wrote:Hi Guys,

Trying to compile on build 765 and getting three errors, see screen shot below...


Any help would be appreciated....
Wiggy

milanese » Thu Feb 05, 2015 8:31 pm wrote:

Wiggy1965 » Thu Feb 05, 2015 10:23 am wrote:Hi Guys,

Trying to compile on build 765 and getting three errors, see screen shot below...


Any help would be appreciated....
Wiggy

just delete int before "c" line 1349, this was coded in this way for using #property strict, without it will give you an error

milanese » Thu Feb 05, 2015 7:27 pm wrote: HGI uses a TMA_Slope no CSS, but right is backtesting makes no sense ..

Cheers

Tommaso

trendinator » Thu Feb 05, 2015 10:37 am wrote:

milanese » Thu Feb 05, 2015 7:27 pm wrote: HGI uses a TMA_Slope no CSS, but right is backtesting makes no sense ..

Cheers

Tommaso

so tma slope cannot be backtested? can you explain why? im just curious is all thanks

milanese » Thu Feb 05, 2015 8:40 pm wrote: because you will got worse results caused by the way TMA is calculated and by the way crappy backtester works.. you will get nothing else than best case results, over optimistic, as signasl are calculating in way that the HGI knows future, what in reality is never the case..

Cheers

Tommaso