Machine learning in trading: theory, models, practice and algo-trading - page 1852

New comment
 

In general terms I see it like this. We have a total number of lines in the file and a variable s that tends from zero to count_size. As a result, in the real trade, the function ReedFile one will run, for example, that stupidly translates the pointer to the line s and reads the data to the end of the file. Thus, when the indicator is called, it will load the exact spot that is missing on the chart. After all, the writing EA will stand on the same terminal as the one that is trading.

I should add that if the delta is calculated correctly and written in the third column, the delta problem will also be solved in a very simple way without overloading the system with data saving when addressing to the fierce cycle of trades for a large period.

Two birds in one solution....

 
Mihail Marchukajtes:

There everywhere you have to specify the current futures contract. It does not work on the splice, of course. And then at once I attach the current version of the indicator which at compilation loads only up to the second bar, thus the first bar always changes, I understand there it is written now the actual OM

I don't have so many instruments in market review. I just wanted to remotely put it there to see how the indicator works.

 
Mihail Marchukajtes:

Alexey, can we make the indicator read the readings from the file for the first bar when a new bar appears and thus it will achieve the most complete synchronization. In fact, it won't stand on the chart, and it will be called periodically from signal to signal... The problem is that the EA is written with the help of Marketbuk and the indicator receives data independently from the exchange. I think we should use market book monitoring and this is what the indicator got. Profit!!!!!

I think the only source of OI should be the EA and the duplication of requests to the exchange is not acceptable, otherwise we just bother to negotiate. What do you think?

I think that the first bar is the second on the chart or it is zero bar?

 
Aleksey Vyazmikin:

Is the first bar the second on the chart or is it a zero bar?

It's the second one on the chart. Not zero.

Remember, this function in the Inite for instruments from 1 to 15 should be mentioned and it will load only C

 if(!MarketBookAdd(Name_instrFS1))
     {
      Print(__FUNCTION__,": "+Name_instrFS1+": Depht of Market data not added!");
      return( INIT_FAILED );
     }
 
Mihail Marchukajtes:

second count. Not zero.

Note this function in the init for instruments from 1 to 15 and it will load only C

So it writes Si Splice_OI.csv file without OI of course.

 
Aleksey Vyazmikin:

So he writes file Si Splice_OI.csv without OI of course.

What do you mean you don't get it?
 
Mihail Marchukajtes:
What do you mean you don't understand?

I deleted everything.

 if(!MarketBookAdd(Name_instrFS1))
     {
      Print(__FUNCTION__,": "+Name_instrFS1+": Depht of Market data not added!");
      return( INIT_FAILED );
     }

left

   if(!MarketBookAdd(Symbol()))
     {
      Print(__FUNCTION__,": "+_Symbol+": Depht of Market data not added!");
      return( INIT_FAILED );
     }

The result is written only file Si Splice_OI.csv

 
Aleksey Vyazmikin:

Removed everything

left

The result is only the file Si Splice_OI.csv

Well, it's uncommented at the bottom for now. Just change Symbol() to 

if(!MarketBookAdd(Name_istrFS))
     {
      Print(__FUNCTION__,": "+_Symbol+": Depht of Market data not added!");
      return( INIT_FAILED );
     }

It just needs to specify the current futures. When you hang the EA on the glue. Splice will not work. If you relying on the market with this symbol, it should work in your account.

 
Mihail Marchukajtes:

Well, it's in and down for now. Just change Symbol() to

It just needs to specify the current futures. When you hang the EA on the glue. Splice will not work. In other words, if you set "Si-9.20", it will work too and you'll be happy that the symbol automatically appears in Marketwatch.

I do not hang on the glue - in general, should sort out. You can not check and send me a working version ;)

 
How do I get the time in microseconds for the first bar? I forgot something...
1...184518461847184818491850185118521853185418551856185718581859...3656
New comment