something is wrong
-
im getting firious with this fxdreema...
I want to know why this ea doesnt make the same on demo as on real....
Just doesnt open the trades at the same levels, on demo it does the things right opening the trades at the correct hihgs lows of the period but on real it calculates it wrong....
I can add that the charts im working at on real are like EURUSDm and not EURUSD.... its the same code for real than for demo and i leave the ea on the vps but it calculated the high and low bad givinig me bad prices and i dont know why because volatilie doesnt affect pretty much it just calculates wrong the prices..
......
AZ 010.mq4 -
Oh boy, what part of this EA is wrong?
-
Something like this - http://prntscr.com/732h8w - is most probably not a good idea for the real world, even for the backtest. This time filter can pass once a day only if there is a tick in one particular second. But what if no tick comes in that second? I will suggest to use wider time period.
The alternative is to use the timer event, which will work on live, but because of technical reasons this event can't be backtested.
-
I did some visual changes here: https://fxdreema.com/shared/dceHpggEd
Something as big as this must be put in visual order. Which is not always an easy task. This is where I lose most of the time in the project called fxDreema, I always make some visual changes in the code
Those blocks Turn ON, Turn Off. It's enough to turn off one block and all the blocks connected after it will not have the chance to run anyway. And it's probably better to give those blocks (to be turned on and off) some good name. Yes, block IDs can be text, not only numbers. But I'm not sure that using these blocks is actually a good idea.
I notice that you like to connect blocks in parallel. This is not always a good idea, because depending on block numbers you can get different results. Look at the very bottom blocks in my version of the project - I connected them one after another and I think this is the better way, it's more clear.
Those Formula blocks... here it's best to calculate vales right before using them. You have a group of Formula blocks in the middle of nowhere... who knows what is going on with these values. It's also a good idea to give good names to their Results. It's not mandatory for the results to be like "Result202", you can give those any valid variable name like "AllowedToSell". Also, you can use the same result name in multiple Formula blocks.
-
I liked the way u organizated the blocks, i dont know why i hadnt done that before.
Issues:
-I have turned off permanently a group of blocks wich were thought to calculate a special lot size but it is too complicated to try it on real because charts on small accs are like EURUSDm and doesnt work so i let them off and setted a new constant for fixed volume (lots) for all pending orders.-About the turn off and turn on blocks thats not the problem because i thought that what u are saying, and i setted them just to turn on/off blocks of buy/sell pending orders wich are always the finall blocks. For that way im sure there cant be a problem.
-About the parallel blocks i also comprehend that and i connect on parallel block wich i dont care wich of them activates first, the only i need, and also is what is exactly doing, is to activet all the blocks that are on the same level (no matter the order) and then starts with the blocks below. So i dont think the problem come from there.
-About the time filter, there i think it would be good to make better that part. I only can imagine this way to do it, and i can say it works on real and demo, but better on demo.
The problems i ahd on real were:
(that didnt happen on demo)
-I had my stops orders well placed with its stop loses and take profits perfectly calculated, the i a sell stop was activated and i think it was at the same minute that the order were closed on sl...thats imposible.. because sl and tp always are almos 50 pips away from open price, so i checked the values on history and it said that the sl and tp wre just 3 pips away!!?? i didnt move them no way, and i also dont remember to have programmed that function on the ea...
The sl and tp were just glued to the open price.. imposible, for good it was only -0.4 usd but the same was that it coulb were a profit of 10usd.-The next problem is that i think that demo data and real data is not the same or something like that because the ea determinates differents highs and lows on the same piriod of time on real and on demo. Just with a difference of 1 or 2 pips may be 3 sometimes but i still would like to know why it is not the same... the last problem was volatilie of markets i think, because the pending order was activated during a new. But my question about this is.. what is correct to happen; that the tp and sl of the pending order displaced mantains the same amount of pips away from open price? or the tp and sl mantains the same values and so the amount of pips away from open price changes??
Still loseing with the ea but i think it will recover, on demo it is doing good results since 45 days with some period of loses and then an other good profits. I will keep it testing on real with this new version well organaized, also tell me if there is a way to replace that time filter blocks... i will backtest it on demoa nd real.
......
AZ 030.mq4 -
For the time filter, the problem is that the backtester runs the EA on every tick, where every tick is generated. Often there are few seconds difference between two ticks. So in the tester you just can't do something on every single second. And this is also pretty much true for live trading as well!
You can be sure that the EA will so something between 12:00:00 and 12:00:01 only if the EA works as a clock, which is only possible with it's Timer event. But again, the Tester does not have such Timer event, it only works with it's generated ticks that appear in irregular time periods.I don't know what is the idea of this EA. Yes, I checked the blocks structure and I tested the EA, but I don't know it's strategy. And as I saw, it actually have 2 different strategies that can be changed with some button. If you give me some small test project where I can see exactly what I need to see... fine. Or you can give me screenshots of what is wrong. But if I have to explore everything in this EA...
-
The ea just calculates the high and low between 0:00 and 12:00 on eurusd, then at 12:01 it places stops orders if we are on Breakout mode or limits orders if we are on range mode.
I use that buttom to change the strategy if i see that one of them is being profitable on the actual period of time, as you can see one strategy is the opposite of the other one, because their orders has the same tp and sl (tp and sl are the amount of pips between the high and low of the period 0:0 to 12:00 and are the same amount tp and sl). Its also not scalpping because i use is on h4). For example actually the breackout mode is getting good trades so the range mode is getting the same trades but as losses. I look to the graph of profits of both strategys ans check the trend of profitablity of each strategy and choose.The fact is that it should work the same as on demo, i dont know may be what is different is because of the broker and because it is on real... i will check on this week if it keeps working strange.
I give you a photo of that sell trade that i dont know why it has closed at this level with -0.4 when it should had been closed with -10 usd or +10 usd......
......
eurusd trade.JPG -
This 2 photos show you the difference of the prices that the ea gives to the pedning orders just a difference of less than 2 pips but i dont know why it make it different of demo, also i show you that the candle data on demo than on live is not the same.
Also tell you that the demo acc almost 45 days before was 500usd and now is 630 usd
......
eurusd1.JPG
eurusd.JPG -
Here is the actual current trade that has opened 10 pips above because the buy pending order was placed wrong by the ea on real but not on demo.
......
eurusd2.JPG -
You have 20 blocks to create orders! This is like a boring game for me now, I only know that I need to find in which chest the treasure is. But my mind prefers to think about something else.
What about some simple EA made out of 4-5 blocks that will show me the problem right after I add it to the chart?
-
Ok take a rest, i will leave it an other week on live and see how it works, it was just the first...
But you have to know that i didnt have any problem like this on demo for that reason im not sure if it is a programing error.May be is that about of time filter and ticks, i dont know one week i will be back with news.
-
This shows Highest and Lowest price level between 00:00 and 12:00 for the current day. If the time if somewhere in between, it shows Highest and Lowest between 00:00 and that time. I want to know what is wrong here:
https://fxdreema.com/shared/ujhC1tQgd
http://prntscr.com/73g96u -
For the Highest and Lowest price, look at the candles on the chart where the EA is working and find the Highest/Lowest price for the given period. The value that the EA calculates must be the same with the value on the chart where the EA is set to work. Not with the value in different symbol in different timeframe in different broker and as what I can see, different time offset.
http://prntscr.com/73gfo0
http://img.prntscr.com/img?url=http://i.imgur.com/iMRE48g.png -
Dear fxdreema about your last comment, thats obvious im surprised you think i missed that .... The screenshot of the left is just the mt4 platform i use on my dekstop to see the acc, there is no ea running, i just see the orders placed on the acc and doesnt matter the timeframe at all, the screenshot of the right is the vps platform where the ea is well placed on h4 as it has to be. The mt4 of the live acc is in the vps in other mt4, but the mt4 im showing you on the left has the same trades.
I mean.. for that reasn its obvious it isnt wrong,
The ea with the blocks with logic point tester is good, but i think there is still the same problem, because that is just another way to calculate the high/low, but then at the time to place the orders there is the problem with the time filters.
