Hi Mr Ed;
Are you still using FXDD as a broker? Do you find it reliable?
Are you still using FXDD as a broker? Do you find it reliable?
Mister ED's 4 hour Heikin Ashi system
- #1,064
- Oct 26, 2009 1:03pm Oct 26, 2009 1:03pm
Thanks longBar!
However, there still seems to be issues that results in a lost where the trigger is between -/+.
2/17/09 04:00 GMT+0 a buy resulted in loss as you can see from my chart
If I am not mistaken your chart seems to have EA v1.5.0 attached - this version does not have the RAVI FX Fish indicator incorporated so filtering would not take place. I guess you would want to upgrade to the EA v1.6.1 for this feature
Longbar
Hi hello,
If I am not mistaken your chart seems to have EA v1.5.0 attached - this version does not have the RAVI FX Fish indicator incorporated so filtering would not take place. I guess you would want to upgrade to the EA v1.6.1 for this feature
Longbar
Sorry for the confusion. I'm trading v1.5 on my live acct. I was backtesting 1.6.1. I compare my visual backtest charts with my real charts and they both showed the same. Someone can backtest around the time I posted and see if they get the same results. Most of my losses should have been filtered by the RAVI filter, but they are very close calls(where the target is on the +/- line. Prob should have something where it is < rather than <= ).
Thank you,
Antoine Walker
- #1,073
- Oct 26, 2009 3:42pm Oct 26, 2009 3:42pm
Hi Longbar,
Thanks for the new EA. I'll be trying it out for the rest of this week. BTW you must be living on a similar time zone to me? GMT +4
Cheers
Close enough - actually I am in Ireland, same time zone as UK
LB
- #1,074
- Oct 26, 2009 4:32pm Oct 26, 2009 4:32pm
Hi Longbar,
Do we need to change the settings on the Ravi filter to 4,24,0.12 when we put it on the charts?
The EA has the correct settings but I wasn't sure if the indicator has to be loaded and set on the chart correctly.
When I load it on the chart the default settings are 4,49,0.09
Many thanks,
Ian
The EA loads all the indicators on its own and those are not visible on the chart. The loaded indicators will be configured according to the EA parameters.
If you want to visually double check what's going on then you can load the (same) indicators on the same chart the EA is attached to. The EA will still use its own indicator instances, independent from the one(s) you attach to the the same chart.
So ensure that the EA parameters for the indicators are set properly (eg HASIndicatorVer, HAS_MAx_xxx, ATRPeriod, RAVIFX_xxx).
If you curious and want to make the indicators visible then backtest the EA in "Visual mode" or click "Open chart" button after a backtest run
LB
- #1,075
- Oct 26, 2009 4:51pm Oct 26, 2009 4:51pm
Hey LongBar,
Sorry for the confusion. I'm trading v1.5 on my live acct. I was backtesting 1.6.1. I compare my visual backtest charts with my real charts and they both showed the same. Someone can backtest around the time I posted and see if they get the same results. Most of my losses should have been filtered by the RAVI filter, but they are very close calls(where the target is on the +/- line. Prob should have something where it is < rather than <= ).
Thank you,
Antoine Walker
In general with MT, 1 pixel on the screen represents a range of values so what may look on the edge could actually be either <, or >, or hmm =
The only way to make a call is to check the Expert log where the IsTradeFiltered function prints the values to, something like this:
2009.10.26 12:55:38 2009.07.01 16:00 MisterED EURUSD,H4: INFO: CheckTrigger - HA colour change found, potential BUY direction trade (BUY,2:SELL)
2009.10.26 12:55:38 2009.07.01 16:00 MisterED EURUSD,H4: INFO: IsTradeFiltered - RAVI FX Fisher filter activated (0.07898968, 0.12000000)
This says the RAVI FX Fisher filter got activated and you see the red and the green numbers in brackets. The red is the actual indicator reading while the green is the threashold (EA parameter RAVIFX_Trigger). When the red figure is in or on the edge of the range of +/- the green one then the filter takes place. This is the exact verification of the EA action to filter or not.
- #1,076
- Oct 26, 2009 5:02pm Oct 26, 2009 5:02pm
Here are the 1.6 results with the RAVI filter. They are alot better!
Instead of doing this
- -RAVIFX_Trigger < Indicator reading < +RAVIFX_Trigger
the expression became this
- -RAVIFX_Trigger <= Indicator reading <= +RAVIFX_Trigger
observe the equal sign.
So if you get different results using the different versions then it is likely down to the RAVIFX_Trigger parameter... of course assuming that you get 90% modelling quality for both runs otherwise the results follow more of a random number generator
----
Edited later:
Just to add, see attached my backtest results over the period of Jan 2007 to Feb 2009 at Alpari UK. I see a significant improvement when the RAVI filter is enabled.
Ohh, that explains
In general with MT, 1 pixel on the screen represents a range of values so what may look on the edge could actually be either <, or >, or hmm =
The only way to make a call is to check the Expert log where the IsTradeFiltered function prints the values to, something like this:
[size=1][color=Blue]2009.10.26 12:55:38...
Longbar,
Found it Can we add some kind of buffer so it will not get triggerend when it is so closed to +- targets. Will probably filter some good trades, but will also stop some of the losing trades. Backtesting with the latest version caused a margin call. 1.6 seems to do ok.
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: CheckTrigger - HA colour change found, potential BUY direction trade (BUY,3:SELL)
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: IsTradeFiltered - RAVI FX Fisher out of filter range (0.12580154, 0.12000000)
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: SubmitOrder - new BUY order at bid:131.58000000 , entry:131.65000000 SL:0.00000000 TP:0.00000000 Lot:0.10000000 exp:1970.01.01 00:00:00 ExitMethod:0 Cmnt:133.0 104.5%
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: open #7 buy 0.10 GBPJPYFXF at 131.65 ok
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: modify #7 buy 0.10 GBPJPYFXF at 131.65 sl: 130.32 tp: 131.68 ok
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: ModifyOrder - successful order SL/TP/exp modification
Found it
Can we add some kind of buffer so it will not get triggerend when it is so closed to +- targets. Will probably filter some good trades, but will also stop some of the losing trades. Backtesting with the latest version caused a margin call. 1.6 seems to do ok.11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: CheckTrigger - HA colour change found, potential BUY direction trade (BUY,3:SELL)
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: IsTradeFiltered - RAVI FX Fisher out of filter range (0.12580154, 0.12000000)
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: SubmitOrder - new BUY order at bid:131.58000000 , entry:131.65000000 SL:0.00000000 TP:0.00000000 Lot:0.10000000 exp:1970.01.01 00:00:00 ExitMethod:0 Cmnt:133.0 104.5%
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: open #7 buy 0.10 GBPJPYFXF at 131.65 ok
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: modify #7 buy 0.10 GBPJPYFXF at 131.65 sl: 130.32 tp: 131.68 ok
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: ModifyOrder - successful order SL/TP/exp modification
- #1,080
- Oct 26, 2009 5:21pm Oct 26, 2009 5:21pm
Longbar,
Found it
2009.10.26 12:55:38 2009.07.01 16:00 MisterED EURUSD,H4: INFO: IsTradeFiltered - RAVI FX Fisher filter activated (0.07898968, 0.12000000)
However when the filter is NOT triggered you will see a different log:
11:22:34 2009.02.17 04:00 MisterED_v1.6.1 GBPJPYFXF,H4: INFO: IsTradeFiltered - RAVI FX Fisher out of filter range (0.12580154, 0.12000000)
If you would like to reserve more room or buffer then I would suggest to change the RAVIFX_Trigger parameter accordingly.
Just comes to my mind the main changes from v1.6 to v1.6.1 was to read the RAVI indicator belonging to the trigger HAS bar (second last) instead of the last bar. This could be another, and probably the main source of the diffs. What do you think?