Indicators development for T101's not very simple method | Page 123 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

SIMPLE EA for a simple method needed!! 2 replies

Simple System By T101 - My Thought 18 replies

Continuation of T101 method discussions 24 replies

My first attempt at T101's basket method... 2 replies

Attachments: Indicators development for T101's not very simple method
Exit Attachments

Indicators development for T101's not very simple method

  • #2,441
  • Jun 16, 2010 9:08am Jun 16, 2010 9:08am
important

with version E, i just hit the stoploss again.

it opened up another order of 12 sells and 2 buys.

5 minutes laters all the signals have all changed from sell into 0 except trend magic. trend magic has now a signal to buy.

now you would expect i would have a lose trade on my hands. not so. i am actually running to into profit now. even if it is only +50 pips at this moment
  • #2,442
  • Jun 16, 2010 9:24am Jun 16, 2010 9:24am
Look guys, I am not as good a coder as Robert nor finimej for sure, and you also have charvo here. I'm sorry for my rants. I apologize, and ask for the more experienced coders to take care of this. I surely thought that almost all of what has been mentioned as problems, are things that I did maybe in a different way, but nothing that would have kept the EA from working. Obviously I am wrong. I am fine with giving what I have to the development of Kos' EA. I really don't want to control the effort when there are more experienced coders than me that can do better code more efficiently obviously.

I can say that on H1 overnight, Kos EA d version, without re-entry rule(s) implemented yet, the EA just was a loser last night (lost 10% - 25% no matter if using profit-locking or not, tested on 4 brokers).

I was thinking of testing Robert's latest uploaded EA, but without the re-entry rules set, and variables made to keep track of number of basket sells (so after first, 5th entry condition would also be used), then I feel it will test as badly as the d version has. If Kos is not getting all these losses, then we haven't mimicked his trading yet. Waiting for a retracement should help, but I still recommend the addition of the two rules I mentioned multiple times earlier, since they just make common sense.

If I wasn't in an overall funk, maybe I'd take up the baton, but since I am, I don't feel like coding or doing much else.

Regards,

Don
  • #2,443
  • Jun 16, 2010 9:54am Jun 16, 2010 9:54am
1) This is not what is wanted. A buy can be done after a buy. It's not a must to go from buy to sell to buy to sell, etc. In fact, Kos never mentioned doing that. He says he waits on a retracement before buying again. If the retracement was mentioned before I coded the EA, I would have added it before. We can do it as I mentioned to Robert, or simply say that this bar's high and low is higher than last bar's high and low, and that last bar's high and low was not both higher than bar before that. That might even be better than re-entry code I mentioned to Robert.

2)Profit-locking function has had an external boolean to turn it off since first EA version I uploaded.

3)zero divide problem has usually been associated with my MM, but it's not a problem in this EA as I see it (I'm getting no zero divide errors, Lots are calculating correctly from what I can see, so I must have remembered to put in all parts of my MM correctly. Usually when I get zero divide, it's either I forgot to put in pip value or that StopLoss is 0). I use an array for pip values, to get pip values for all pairs. Is there something wrong with it? Have you seen it not work correctly?

4) That's a lot of code. Looks like a good place to get code from. Honestly overwhelming at the moment.

Regards,

Don


Quoting finimej
Don,
I have read through the code roughtly. Here are some suggestions. It is better you correct the code by yourself, so that you can track the changes of the code.

1) Logic part.
Current your code, if all 4 conditions meet for buy operation, then systtem does buy. If the lock profits exited the buy operation, the all 4 conditions still meet for buy operation, then system entried the buy again. This can be fixed use the above switch.

If we do not do scale in, then first we need to check the number of open positions. //the system current...
  • #2,444
  • Jun 16, 2010 10:35am Jun 16, 2010 10:35am
Close trade with -560 pips as SL = correct. cos my trade just hit SL now after +350 pips

version D, 1hr TF, etc

Therefore the SL is working properly
Also the MACD close is working properly

kos
John 3:16 - From THIS love no power on earth can me sever.
  • #2,445
  • Edited 1:12pm Jun 16, 2010 11:06am | Edited 1:12pm
OK, I figure the biggest difference between what Kos sees and what the rest of us see is due to Trend defined by Weekly and Daily, and the difference is due really to GMT Offset that Kos' broker uses.

ATC brokers and MBT are both GMT -4.
IBFX is GMT +0.
CitiFX is GMT+1.
Alpari UK is GMT +2.

Because of the above differences, sometimes there is hours between when Kos gets a signal and when the rest of us do. Because of this, I recommend GMT Offset code be implemented. If not, then signals will differ between all of us even if using all same other 3 indicators. Most of the code to implement is in GMT Offset indicator I uploaded recently. I have done work on it since then in another EA a while back, so maybe I can dig up the specific code to make it easy to implement as possible. Kos' GMT Offset is 2 (GMT+2) at Alpari UK, so that would be the stake in the ground as far as Weekly and Daily open, then using GMT Offset, we get everyone else on the "same page".

Regards,

Don
  • #2,446
  • Jun 16, 2010 2:57pm Jun 16, 2010 2:57pm
Hi Don!

I dont know if this has a way of affecting the trades, though I know there are some EAs that base there time to look for set up on time and those kind of EAs use GMT to synchronize when the set up happens across different brokers

Anyway anything that will make the EA functional but which will not take the SIMPLICITY of the strategy away is welcome

and welldone bro, for the hardwork

kos
John 3:16 - From THIS love no power on earth can me sever.
  • #2,447
  • Jun 16, 2010 3:26pm Jun 16, 2010 3:26pm
Quoting disbellj
OK, I figure the biggest difference between what Kos sees and what the rest of us see is due to Trend defined by Weekly and Daily, and the difference is due really to GMT Offset that Kos' broker uses.
Use PERIOD_H1 data to determine the starting price for the Day and/or Week, not PERIOD_D1 and PERIOD_W1.

As I did with the Basket script, you can assume each day starts at 17:00 New York time and adjust each broker's starting time accordingly. Or designate another time as the "official" start of each trading day - maybe something like 00:00 CET. Just make sure you take Daylight Saving Time into account when converting times between timezones.
  • #2,448
  • Jun 16, 2010 3:30pm Jun 16, 2010 3:30pm
Trades happen in realtime right?

so whatever Kos or me or you have on the chart is whateverbody else has.

the only difference i the time setting on the chart itself since we all live in different parts of the world.

the only difference today was that Kos jumped in early in the game. he has his Ea running ( D version ) when the 4 signals were not lined up yet.

when i jumped in with the E version. the 4 signals were already lined up.

i dont think that has something to do with GMT time or ofset. but more at when we had applied the EA on the chart. he was earlier then me, so we have different results.

in the end though i also went 500+ pips down. so that is the same.
  • #2,449
  • Jun 16, 2010 3:34pm Jun 16, 2010 3:34pm
zz,

Yes, it looks like Kos' broker uses same GMT offset as FXDD, which goes by the start/end of the Forex day 17:00. I have DST code too (which basically shifts one hour if DST is on, which I use a user input, just like I do for GMT Offset). zz it looks like you must figure all that out internally since I don't see user inputs for GMT Offset or DST.

Do you agree that it would be a good idea to make these the same between brokers, so we mimic/see what Kos' does (as close as possible)? I think so.

You obviously have more advanced code than I do to take care of this problem. Do you mind looking at latest EA, at buy and sell condition number 4?

Other than that and retracement code (as I told Kos about, which also takes higher high into account), and adding that this bar's price must be above TM (not just last bar's), I don't know of much else that could be done. I think after those changes made, we'd be as close to Kos' strategy as he gave it to us, taking everything into account I can think of.

Regards,

Don


Quoting zznbrm
Use PERIOD_H1 data to determine the starting price for the Day and/or Week, not PERIOD_D1 and PERIOD_W1.

As I did with the Basket script, you can assume each day starts at 17:00 New York time and adjust each broker's starting time accordingly. Or designate another time as the "official" start of each trading day - maybe something like 00:00 CET. Just make sure you take Daylight Saving Time into account when converting times between timezones.
  • #2,450
  • Jun 16, 2010 3:37pm Jun 16, 2010 3:37pm
trade4fun,

I highly disagree. Because each broker has different open times for week and day, what Kos is looking at is NOT what we're looking at (unless you happen to be using Kos' broker or FXDD or another broker using GMT+2).

Maybe, just maybe, Kos jumped in when his signals lined up. Maybe, just maybe, your signals hadn't lined up yet for the exact reason I've stated - GMT Offset of the broker being different than Kos' broker. Just because today the numbers aren't VERY different, doesn't mean that most of the times they are not. Today they were closer between brokers than usual. Why? The ranging lately is the obvious reason IMHO.

The other 3 conditions may look the same on every broker, but I'm concerned right now with the 4th condition, which is where GMT Offset would come in, and there is no doubt in my mind that having GMT Offset implemented and setting open time for week or day to 17:00 (same as Kos' broker's and FXDD's open/close), would put us all on the same "trend" start, so we would all be on the same page as close as possible. Having said that, those that do not use IBFX pair-set would have different trend numbers than those that do use IBFX pair-set (and we can't get around that other than use IBFX pair-set like Kos does, which is default in EA).

Regards,

Don


Quoting trade4fun
Trades happen in realtime right?

so whatever Kos or me or you have on the chart is whateverbody else has.

the only difference i the time setting on the chart itself since we all live in different parts of the world.

the only difference today was that Kos jumped in early in the game. he has his Ea running ( D version ) when the 4 signals were not lined up yet.

when i jumped in with the E version. the 4 signals were already lined up.

i dont think that has something to do with GMT time or ofset. but more at when we had applied the...
  • #2,451
  • Jun 16, 2010 3:44pm Jun 16, 2010 3:44pm
Quoting trade4fun
Trades happen in realtime right?

so whatever Kos or me or you have on the chart is whateverbody else has.

the only difference i the time setting on the chart itself since we all live in different parts of the world.

the only difference today was that Kos jumped in early in the game. he has his Ea running ( D version ) when the 4 signals were not lined up yet.

when i jumped in with the E version. the 4 signals were already lined up.

i dont think that has something to do with GMT time or ofset. but more at when we had applied the EA on the...
trade4fun

YOU ARE WRONG, I never jumped in prematurely, read some posts back. I showed when the 3 conditions are 1 and I was waiting for close below TM and as soon as the bar closed below TM, THE EA TOOK THE TRADE. and I posted EA entry chart, and left it to do its thing till the end.

If your EA did not take that trade then Don is right about the GMT talk OR you are wrong about my entry

kos
John 3:16 - From THIS love no power on earth can me sever.
  • #2,452
  • Jun 16, 2010 4:00pm Jun 16, 2010 4:00pm
@ Don

pls look at the orange circles in my pix below

dont you think the sellcondition4 should be 1 now since samir daily has gone short strong. or is the calculation in the code different from what samir does

as for BI_T101 I zoomed in on it and I realised the cross has not really happened

kos
Attached Image (click to enlarge)
Click to Enlarge Name: short missed.gif Size: 42 KB
John 3:16 - From THIS love no power on earth can me sever.
  • #2,453
  • Jun 16, 2010 4:16pm Jun 16, 2010 4:16pm
@ trade4fun

as the bar closed all 4 conditions lined up (look at the green circles) and the EA jump in to place the trade NOT ME jumping in as you claimed

see the chart below

kos
Attached Image (click to enlarge)
Click to Enlarge Name: short missed taken.gif Size: 46 KB
John 3:16 - From THIS love no power on earth can me sever.
  • #2,454
  • Jun 16, 2010 9:11pm Jun 16, 2010 9:11pm
Kos,

I have told multiple times on the problems with Samir's indicator (wrong bar used is main problem since MT4 gives you wrong bar as start bar for week open, maybe day open too, and I THINK either spread or currency value not taken into account. Main problem definitely was MT4 startbar given when ibarshift used with Week Open time. So I have corrected it to work on the fly. I trust mine, since it's been checked to real prices and agrees).

However, in your post, your pic shows both Short Strong. Why was the trade not taken? BI is red over blue, for buy. Until it crossed, no sell trade would be taken, per your rules. All 4 have to line up. It doesn't really matter which is first, since all have to be aligned for a trade to be taken.

Regards,

Don


Quoting Kosomolate
@ Don

pls look at the orange circles in my pix below

dont you think the sellcondition4 should be 1 now since samir daily has gone short strong. or is the calculation in the code different from what samir does

as for BI_T101 I zoomed in on it and I realised the cross has not really happened

kos
  • #2,455
  • Jun 16, 2010 9:14pm Jun 16, 2010 9:14pm
There ya go. Like clockwork, just as it was coded to do (of course, that was Kos' vision, I coded, and Robert fixed the actual trading functions, so it's been a group effort to get it done. I of course thank Kos and Robert both for their generosity).

One edit to Kos' pic: He outlined Total on Samir's indicator, when really Kos looks at Long Strong or Short Strong, which is not taken from Total, but from Buys and Sells values individually and compared to each other (counting spread and currency value if one mini lot per pair, and using correct start bar for day and week, at least on mine).

Regards,

Don


Quoting Kosomolate
@ trade4fun

as the bar closed all 4 conditions lined up (look at the green circles) and the EA jump in to place the trade NOT ME jumping in as you claimed

see the chart below

kos
  • #2,456
  • Jun 17, 2010 12:13am Jun 17, 2010 12:13am
Here is the trade in progress, I dont know if any one gets this trade or am I the only one testing this work. If you are on it let us compare notes.

its +440 pips now. 1hr TF/ no locking / rest default/ version D

Now I can enable locking to protect some of the profit, though this is demo, but I have to treat it like live!

NB - 3 CONDITIONS ALIGNED IN 4HRS TF WAITING FOR SAMIR WEEKLY if that happens then we have a long ride to take

kos
Attached Image (click to enlarge)
Click to Enlarge Name: short missed taken on.gif Size: 45 KB
John 3:16 - From THIS love no power on earth can me sever.
  • #2,457
  • Jun 17, 2010 1:04am Jun 17, 2010 1:04am
A lot of lurkers on this thread, big names at that. I guess they are waiting to pick this finished toy, or compiling the rules to write their own secret toys. pls dont leave the testing to only 2-3 of us.

look at 4hrs set up, the RED circles are OK and the BLUE circle is keeping us out. Now 3 sell pairs have moved up by the time 4 or 5 move up may be we will be ok to go. So that you understand what your EA is doing and in case you want to do it manually.

We will see what happens later.

kos
Attached Image (click to enlarge)
Click to Enlarge Name: 4hrs set up.gif Size: 32 KB
John 3:16 - From THIS love no power on earth can me sever.
  • #2,458
  • Jun 17, 2010 1:09am Jun 17, 2010 1:09am
hi Kos

you are not the only one with this trade...i closed the trades with profit this morning manualy, because the price goes sideway now...but i forgot to deactivate the ea before, so it opened the sells again......now i'm waiting to see if the trend goes up or not ( the old signals from sells are still there)...then i will activate the ea (version D) again...

as i mentioned before, we get more information from the chart with our eyes then an indi can...so i watch what the chart shows...then i swich the ea on or off, when signals appear in the one or the other direction...i'm lazy to open 14 trades at time... this is a manual trading, not a "set and forget" thinggy...

Thanks MrPip for the corrected version...

have a nice day or sweet dreams guys..
Cado
  • #2,459
  • Jun 17, 2010 1:10am Jun 17, 2010 1:10am
tz...you are faster then me, Kos...

i took the H1 trades on GBPJPY...
  • #2,460
  • Jun 17, 2010 1:39am Jun 17, 2010 1:39am
Quoting Kosomolate
A lot of lurkers on this thread, big names at that. I guess they are waiting to pick this finished toy, or compiling the rules to write their own secret toys. pls dont leave the testing to only 2-3 of us.

look at 4hrs set up, the RED circles are OK and the BLUE circle is keeping us out. Now 3 sell pairs have moved up by the time 4 or 5 move up may be we will be ok to go. So that you understand what your EA is doing and in case you want to do it manually.

We will see what happens later.

kos...
Hi ,

I would like to help test, I am using IBFX demo , what timeframe and version would you like me to test ?
2 traders viewing now