Libraries: Report - page 15

New comment
 

Forum on trading, automated trading systems and testing trading strategies

Libraries: Report

fxsaber, 2023.11.30 12:43 pm.

For market orders slippage can only be determined by analysing tick history. This has not been implemented yet.

Implementation.

Расчет проскальзываний маркет-ордеров.
Расчет проскальзываний маркет-ордеров.
  • 2025.02.18
  • www.mql5.com
В MT5 маркет-ордера не хранят цену, по которой был сделан запрос на совершение сделки. Вычисление исходной цены маркет-ордера. Однако, это значение возможно получить через тиковую историю, которая
 
Now it's easier to identify martins/greeders.
Нестандартный анализ истории торговли.
Нестандартный анализ истории торговли.
  • www.mql5.com
После того, как ТС прошла массу проверок на бэктестах/демо, приходит время реальной торговли. Эта логика порождена двумя гипотезами: Торговля на реальном счете и затем прогон на истории покажут
 
fxsaber #:
Now it's easier to identify martins/greeders.

A strange hypothesis has been formed.


The probability of fitting during optimisation decreases if you optimise the original EA with the grid mechanism enabled.

Roughly speaking, the custom optimisation criterion for the original EA is the same criterion but for EA+Grid.

 

Вероятность подгонки во время оптимизации уменьшается, если оптимизировать исходную ТС с включенным сеточным механизмом.

Roughly speaking, the custom optimisation criterion of the original EA is the same criterion but for EA+Grid.

Can you chew it up a bit? I smell something interesting, but I don't get it yet ))))
I would be grateful!

 
Sergey Porphiryev #:

Can I chew on that a little bit?

For example, there is a TS with no fills - no more than one open position. And we need to set it up through the optimiser.


How it is usually done.

  1. You run it through the optimiser on some custom criteria.
  2. The best variants were examined for good trading on unknown data.
  3. If it's good, we run it in real life. And then - as it goes.

And it can (almost always) go badly.


So here is a hypothesis that we can probabilistically reduce this "shit" if we modify the custom criterion on EA+Grid.

Before the above three items, we allow the TS to take a share according to the grid principle with a lotness coefficient of one.

Further, all three points are fulfilled, but we run without fractions in real life.

 
fxsaber #:

A strange hypothesis was formed.


The probability of fitting during optimisation decreases if the initial TC is optimised with the grid mechanism switched on.

Roughly speaking, the custom optimisation criterion for the original EA is the same criterion but for EA+Grid.

Why is it weird? it works as a regulariser, it adds a penalty for model complexity (less accurate trades, less overoptimisation).

But then, if you remove the grid, trades can be less accurate.
 

Fixed.

 

Hi, do you perhaps have a version that does not show the slippage () in the profit column in the detailed report or is there a way to not show these slippage values

 
Alexander Jeffrey Klein #:

Hi, perhaps you have a version that does not show the slippage () in the profit column in the detailed report, or is there a way to not show these slippage values?

It's easy to do, but just why?

 
fxsaber # :

It's easy to do, but why?

I want to export the report to excel where I then can study by specific weeks, days and hours

For now the profit column gets imported as 2 values in one column in excel. Maybe easier if a separator like ";" is added to the report then excel can split the values when importing?

1...8910111213141516
New comment