How to convert datas for backtesting in MT4

New comment
 

Hi to all,

I've made a .csv file with 18 120 000 rows for the total year 2011.

Each row is : Date(DD/MM/YYYY) ; Time(HH:MM:SS) ; TickBid ; TickAsk   (separator field is ';' and row separator is CR).

Does somebody could indicate me the right way to transform those datas in a compatible format to backtest with MT4,

and also, if my backtest will take all the year 2011 in consideration ?

Thanks for replies.

LeTigre30

 
LeTigre30:

Hi to all,

I've made a .csv file with 18 120 000 rows for the total year 2011.

Each row is : Date(DD/MM/YYYY) ; Time(HH:MM:SS) ; TickBid ; TickAsk   (separator field is ';' and row separator is CR).

Does somebody could indicate me the right way to transform those datas in a compatible format to backtest with MT4,

and also, if my backtest will take all the year 2011 in consideration ?

MT4 does not use tick data for the Strategy Tester . . .  it makes it's own synthetic tick data from the M1 data . . .  so if you want to use your data you first need to convert it to M1 and the other timeframes that you wish to use in the Strategy Tester.  The csv2fxt  script will do this for you as long as your tick data is formatted correctly.  Use Google to search for csv2fxt . . . 
 
RaptorUK:
MT4 does not use tick data for the Strategy Tester . . .  it makes it's own synthetic tick data from the M1 data . . .  so if you want to use your data you first need to convert it to M1 and the other timeframes that you wish to use in the Strategy Tester.  The csv2fxt  script will do this for you as long as your tick data is formatted correctly.  Use Google to search for csv2fxt . . . 


Hi Raptor,

Thank for the info. I've done this and found the Simple_CSV2FXT indicator, and have also downloaded the FXTHeader in folder ..\Experts\include\.

I've also dowloaded an independant broker's version of MT4, to test my own datas, but the setup is blocked when the green bar is full ? I don(t know for why ?

Concerning the backtest, do you think my tick datas are not better precise than the generated ones by the tester ? Can we be secure whith the synthetic data from m1 datas ?

If I give the total 2011 year m1 data to the tester, does it takes all datas in consideration ? and how many ticks it generates for 1 minute ?


To conclude on the same way, do you know where I can find a tuto to record datas in real time for indicators which are set on a chart and on charts below the main one ? it is for making a correlation study between 5 indicators.


Thanks,

New comment
Reason: