Gathering a team to develop an IO (decision tree/forest) in relation to trend strategies - page 20
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The person who develops the trading class for me said that something is not working out correctly there compared to the demo account from Otkritie.
I'll be honest with you, I haven't done any research on this myself, I use information provided by various sources I trust.
Ask this man what he means.
Regarding delays on MQ-Demo. What difference does it make to the test if there is a delay or no delay?
Check with this person to see what he means.
Regarding the delays on the MQ-Demo. What difference does it make to the test if there is a delay or no delay?
There's something to do with Id position or transaction... Do you really need it?
About the delay - if there is a good result, it will be explained by this delay. And are you sure there is no rubbish in the quotes?
There's something about Id positions or deals... Do you really need it?
Pure curiosity. Maybe you might need to test something someday. But right now it seems like you do...
And you're sure there's no rubbish in the quotes?
No, I'm not sure, but looking at the realtime market I don't see any difference (it wasn't yesterday though). The best way is to open the demo and compare. And the monitoring program for comparison of deals is written quickly enough. All you need is a desire.
Pure curiosity. I might need to test something someday. But right now you kind of need to...
One of the problems is that in the demo and in the real the price of the deal will be different, because in the demo we open at the current bid or ask, while in the real the actual price is not defined before the deal is made in the market, and there may be many deals per trade. And this feature affected the trading functions - I don't know the details. There was something else, but this is 2017, I think these are standard features that anyone who writes code for real market has to deal with. Oh, and plus there were problems with clearing, I remember...
No, not sure exactly, but realtime looking at the stack didn't see a difference (though that was far from yesterday). The best way is to open the demo and compare. And the monitoring program for comparison of deals is written quickly enough. All we need is a desire.
Of course, I can do a lot of things myself, maybe you have already done it, that is why I am asking.
Sounds like you're going to hit the market with a lot of volume in the tests... Okay. И... you want to run the test immediately on the real... ok. My job was just to suggest it.
Sounds like you're going to hit the market with a lot of volume in the tests... ok. И... you want to run the test immediately on the real... ok. My job was just to suggest it.
I'm not going to use large volumes in the tests, if I think the system may flounder, I will open a signal on the real, and if not, I won't.
If I think that it works, I want to use it for real and if it doesn't work - what for?
The intermediate class that works with orders is well developed and functions work properly there, so there is nothing much to check on a demo. But it would be interesting to assess how everything will work on the real account, including the calculation speed of all those predictors... And then compare the result with the tester and analyse the discrepancy.
All right, good luck in your endeavours.
All right, good luck in your endeavours.
Thank you. (chuckles)
A class for mt4 is available to switch trading to:
Real - All trades go as normal.
Real with virtual stops - Trades run as normal, stop levels are monitored and stored locally only.
Virtual Trading - All trades are executed locally. Statistics are counted and stored. Plus virtual trades are shown on the chart.
The class allows virtual trading on a real market. And allows you to switch the mode without changing code in the Expert Advisor.
This is interesting information.
Alexei, the fact that you are thinking about Moech is very good. But I don't see any problem. I have opened a real account in opener and I get real quotes without any hassle on MT5, so I have practically transferred my TS there. The interesting thing about Moeh is that the information about trades is available there, like the market depth and Open Interest, etc. The main thing in the AI is the maximum diversity of incoming data from the market. We are already looking for useful predictors and then we build a model on them.
Let me immediately say that the most interesting and volatile instrument is the Si (non-cash dollar versus the ruble). During the day it moves approximately as the FUNT. For other instruments the picture is sluggish. Volatility is much lower than forex ones.
I have got everything in MT5 because of data collection in indices, but they have single threaded and it slows down the terminal. I cannot use it in tester and I don't have enough capital to start with. If I do not have enough capital I may start from scratch. To Moech.... my homeland :-)