Control Points

 

If i backtest by using the control points,aka "the previous smaller timeframe's ticks", does that include by default the Open and Close price of the bar?

So for example, if we have a bar on H1, if i backtest with control point, then will the open and close price of the bar will be simulated, the in between ticks i dont really care only the open and close value, to check wether that bar was bear or bull or neutral.

Also what if the bar is a doji,or for example,a spinning top where the open and close are very near/or the same.Given the case where they are very near, since the CP method is a bit crude, will it simulate here the Open and Close price separately too?

 
Proximus:

If i backtest by using the control points,aka "the previous smaller timeframe's ticks", does that include by default the Open and Close price of the bar?

So for example, if we have a bar on H1, if i backtest with control point, then will the open and close price of the bar will be simulated, the in between ticks i dont really care only the open and close value, to check wether that bar was bear or bull or neutral.

Also what if the bar is a doji,or for example,a spinning top where the open and close are very near/or the same.Given the case where they are very near, since the CP method is a bit crude, will it simulate here the Open and Close price separately too?

Control points are not well documented .... in terms of how they work. You'll be better off asking metaQuotes directly for that information.

Instead of asking you why you'll choose to use control points, these links might help.

https://www.mql5.com/en/forum/142671

https://www.mql5.com/en/articles/1511

 
ubzen:

Control points are not well documented .... in terms of how they work. You'll be better off asking metaQuotes directly for that information.

Instead of asking you why you'll choose to use control points, these links might help.

https://www.mql5.com/en/forum/142671

https://www.mql5.com/en/articles/1511


Well my EA eats alot of memory, and since described that i need only open and close points, i thought that it might be a good idea to use control points.
 
Proximus: Well my EA eats alot of memory, and since described that i need only open and close points, i thought that it might be a good idea to use control points.

Nah, control points are the worse... maybe because no-one understands it :)

The variable_memory_space your program takes up doesn't change with different tester modes.

If your program is slow during back_testing there are a couple of options you can use.

Each option have its distant advantage or dis-advantage.

1) Optimize the codes.

2) Use OHLC testing modes.

3) Use Open_Only.

4) Code functions like Once_Per_Bar.

Your have to keep reading the forum or learn by experience what advantages or dis-advantages these offer.

 
ubzen:

Nah, control points are the worse... maybe because no-one understands it :)

The variable_memory_space your program takes up doesn't change with different tester modes.

If your program is slow during back_testing there are a couple of options you can use.

Each option have its distant advantage or dis-advantage.

1) Optimize the codes.

2) Use OHLC testing modes.

3) Use Open_Only.

4) Code functions like Once_Per_Bar.

Your have to keep reading the forum or learn by experience what advantages or dis-advantages these offer.

Yes but this EA is not an ordinary one.I just finished a backtest with tick values and then control points, and it gave the the exact same results in 5 different pairs.So since i didnt used only reference to open/close price, i think it will hold.If my EA would need high/low prices for computing then it would be a problem, but not here.So ty anyway for clarification.
Reason: