@anse33 Square brackets refer to individual array elements. For example, array[1] will return the second element for an array, because the indexing starts at 0.
Now, if there are no open orders, OrdersTotal() will be 0.
This:
[OrdersTotal()-1]
will return value of -1. Arrays don't have negative indexes - its out of array range.
Try adding a trade count filter before the block in question.
@cesardefez just do it like in the upper picture, its fine.
Hi, I am not miro, just another bored guy, but
try setting the "A: this block can create 1 object" to "no".
@sidmcfarland you have a example EA proving this bug? Maybe draw a vertical line every time the MAs cross?
@jsauter86 just a basic rule in statistics. More data means more proof. Less data means less proof.
If you cant have more trades in a year, then you should test multiple years.
I would love to see this same backtest result, but started 5 years earlier.
@migen
https://fxdreema.com/shared/BNcQoJLzb
This is how I would do it.. Hope you are comfortable with variables 
@sidmcfarland yeah, maybe that is the bug inside the cross operator: it forgets to check the = case, just like I did
@jsauter86 pretty good! How about 2013-2021 ?
2014 and 2017 saw some big trends on eurusd, that would really stress test the ea
@migen
To build on my previous example, I would do it like this:
The 4 conditions now pass twice (for 2 timeframes) so the total "score" for buy and sell can be 8. I also added a block after the buy > sell comparison, where you can define how big the bigger "score" should be.
This doesn't exactly separate between timeframes, but I think it's close enought to your needs and you surely can modify it from there.
https://fxdreema.com/shared/h52NKD4te
https://fxdreema.com/shared/dCBq6vAYc
As a test, here's an EA which draws 3 vertical lines to candles 1, 2 and 3.
However, due to using an array to name the objects (I think), the lines and names dont match, something goes wrong..
@fxDreema would you please look at this, or explain where I made an error
@jsauter86 for me, recovery factor optimizing works better than just balance, but that might depend on the strategy
@rafaelgrecco said in How to Select "Least profitable trade"?:
What I was missing was:
- Reseting the variable "smallestprofit" on every new loop
We must reset the variables at some point, so the ea don't get "stuck" to some big loss that is already closed. Now it resets the variable on every tick, and also calculates it again so we always have the correct value. It is not the most cpu-optimal way but should work fine.
- Using the second output of "Each Trade" to wait for all trades to be checked
I think this is the correct way, we must first check all trades in order to find the lowest profit.
The looping blocks work a bit different than other blocks, go ahead and test different logic builds 
@tantan thats the way to resolve things
@jsauter86 I mean the optimization algorithm in the tester, you use that?
Recovery factor = net profit / biggest drawdown
Mine works differently.
Here, take this .mq4 and import it to fxdreema so you get the exact same project. I also added a trade count condition for comfort.
0_1531746692796_hedge.mq4
When viewing the generated code, the OnInit() function always contains the code shown below.
What is the purpose of it? It seems the code stops the initializing process if the init reason is a change of symbol or change of timeframe.
Well that seems appropriate for most cases, but why the else statement? Now the init is stopped in any case, right?
The problem: Lets say a project contains some graphical objects on the OnInit() tab, and some boolean constant input determines if those objects should be drawn or not.
If the user changes this boolean while the EA is running, the changes will not take effect because the OnInit() is stopped before those blocks are executed.
I often delete that else statement manually to make those objects inputs work properly, but I'd rather have this code changed for good.
Example: https://fxdreema.com/shared/pd9JF900c
@jsauter86 the genetic algo finds the best inputs using backtest information between 2009-2015. After that, it tests if those inputs hold in the "future", 2015-2020. If forward test fails, we know the good backtest was just luck ("overfitting").
https://en.wikipedia.org/wiki/Training,_validation,_and_test_sets
When you find yourself copying the same block over and over again, there is usually a way to loop the process:
This method creates as many trades as you want, and assigns a special number for each one.
That number (group #) can then be used to control the trades.
Doesn't really answer to your question though, just wanted to give another way of doing it
Maybe some buttons accidentally use the same name?
@cesardefez add these blocks to the system. It forces the buttons to agree with the variables.
https://fxdreema.com/shared/Bto5C2CYb