Machine learning in trading: theory, models, practice and algo-trading - page 3046

 
mytarmailS #:

I think so

How? :)

 
Maxim Dmitrievsky #:
Autoclicker needs to be written

Haha - maybe :)

 
Aleksey Vyazmikin #:

How? :)

I don't remember, try pressing "no" and see if packages are installed.

 
Aleksey Nikolayev #:

You got it right, compare the two trading stats for clarity:

thanks

 
mytarmailS #:

I don't remember, try pressing "no" and see if packages will be installed

In general, all this is nothing - wrote errors all round, tried directly without studio

> update.packages(ask='graphics',checkBuilt=TRUE)
--- Пожалуйста, выберите зеркало CRAN для использования в этой сессии ---
Предупреждение: недоступен индекс для хранилища https://mirror.cedia.org.ec/CRAN/src/contrib:
  не могу открыть URL 'https://mirror.cedia.org.ec/CRAN/src/contrib/PACKAGES'
> 

Tried different storage.

Apparently sanctions.

 
Aleksey Vyazmikin #:

In general all this is nothing - wrote errors in a circle, tried directly without studio

Tried different storages.

Apparently sanctions.

There are no sanctions. There are no R issues.

But R doesn't like a mishmash of versions. If you mix it up and something goes wrong, you can't find any ends.

Start from the beginning: install R from scratch, and then to this version packages without any RStudio. I take the cloud mirror. You have to be careful with packages that are not in CRAN mirrors. And don't get any of your own stuff.

I regularly update R to the latest version (now 4.2.3 is 4.3.0), which is not necessary, but so convenient. I've come across packages that are not in newer versions, and packages from older versions are impossible to install. But always managed to find a more modern analogue - R is monstrously redundant.
 
СанСаныч Фоменко #:

There are no sanctions. There are no R issues.

But R doesn't like a mishmash of versions. If you mix it up and something goes wrong, you can't find any ends.

Start from the beginning: install R from scratch, and then to this version packages without any RStudio. I'll take the cloud mirror. You have to be careful with packages that are not in CRAN mirrors. And don't do any of your own stuff.

I need different versions.

It says "Warning: index for storage is unavailable" - what kind of index - it's not clear at all. I checked the link directly in the browser - some file is downloading.
 

Has anyone understood thequanstrat package for creating, backtesting strategies of any level, optimising parameters and so on?

Created by real traders from the fund and is used every day for strategies with real money.

brief introduction

Some interesting thoughts from the same place.

Instead of using backtests to validate good trading strategies, I think they are better suited to reject strategies that we definitely do NOT want to use.

====

How do I know that my strategy is not over-leveraged or false in its returns?

  • For financial data, upsampling or resampling can be used (but this needs to be done very carefully because of regime shifts).
  • Parameter sensitivity/stability testing. Basically, if I decide to use different parameters for my technical indicators (for example, SMA-30 and SMA-180) or start on different days of the week, will my results change a lot (how fragile are my indicators). model based on its initial conditions)?
 
Aleksey Vyazmikin #:

I need different versions.

It says "Warning: index for storage is unavailable" - what kind of index - it is not clear at all. I checked the link directly in the browser - some file is downloading.

As usual: You need something you can't have, kind of called S&M.

 
СанСаныч Фоменко #:

As usual: You want what you can't have, sort of called S&M.

Why can't you, if it is provided even in R-Studio.

The trouble is that libraries are no longer supported and they don't work in new versions - that's why you need to keep different versions.