Time to convert libraries to MQL5 - page 10
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
It didn't please you. Compare the composition of R and the composition of these packages.
"R" (as well as "S") is a programming language, not a library !
"R" (as well as "S") is a programming language, not a library !
R is a language and environment for statistical calculations and graphics. It has about 3500(!) packages assembled into so-called libraries (not to be confused with dlls).
Above I wrote about five groups of packages of interest to us (that's besides the basic tools). Here's a link to time series capabilities.
Here's a link for statistics.
This is a professional package on statistics, particularly its applications to economics.
It has a huge number of publications - textbooks, monographs - which explain the application of packages in R.
Your opinion is completely out of step with reality. The R language itself does not deserve our attention. I don't see its practical superiority to MQL, although it seems to be more powerful (or maybe not - it doesn't matter). But its value is in packages and literature to those packages.
R is a language and environment for statistical calculations and graphics. It has about 3500(!) packages, assembled into so-called libraries (not to be confused with dlls).
You say a lot of right things about R, but you see wrongly its advantage compared to ALGLIB/FANN.
The main advantage of R as a language and development environment is that it allows a huge productivity in prototyping trading systems. For that time that amateurs of programming spend on binding a couple dozens of ALGLIB methods, you can perform several dozens of experiments in R (equity calculation based on buy/sell signals with allowance for spreads and commissions is made literally in five lines).
In addition, to speed up slow calculations, there are very simple and easy to use parallel programming tools.
R is much slower than C++/MQL5, so the final version of the strategy should be written in something else.
Porting R or its libraries to MQL5 makes no sense, but the MQL5-R interface could be useful. If I have to decide what to bundle it with (R/MATLAB/MATHCAD), I'd definitely go for R.
R is a language and environment for statistical calculations and graphics. It has about 3500(!) packages, assembled into so-called libraries (not to be confused with dlls).
Above I wrote about five groups of packages of interest to us (that's besides the basic tools). Here's a link to time series capabilities.
Here's a link for statistics.
This is a professional package on statistics, particularly its applications to economics.
It has a huge number of publications - textbooks, monographs - that explain the application of packages in R.
Your opinion is completely out of step with reality. The R language itself does not deserve our attention. I don't see its practical superiority to MQL, although it seems to be more powerful (or maybe not - it doesn't matter). But its value is in packages and literature to those packages.
I downloaded and looked through R package, I haven't found something that wasn't in ALGLIB.
Surely I am wrong in your opinion, so point the finger, what is in R and not in ALGLIB?
The reason for porting R or its libraries to MQL5 is absurd, but the MQL5-R interface could be useful. If I have to choose what to bundle it with (R/MATLAB/MATHCAD) - I'm definitely in favour of R.
I have a connection for MQL4. It works. I checked it myself. I have no problems at all with porting to MQL5.
So forget about R and offer me something that can be ported.
I have downloaded and looked through the R package, I can't find anything in ALGLIB that isn't there.
Surely I am wrong in your opinion, so point the finger, what is in R and not in ALGLIB?
I've only seen the table of contents of ALGLIB - it's ridiculous, compared to R. I don't want to dig and look for differences, and don't want to convince anyone I'm right.
R is a specialized statistical package.
PS: haven't seen ARMA, ARCH? just for the record.
I have downloaded and looked through the R package, I can't find anything in ALGLIB that isn't there.
I must be wrong in your opinion, so point the finger, what is there in R and not in ALGLIB?
For example, let's look at hypothesis testing... http://alglib.sources.ru/hypothesistesting/
Not a single unit root test. Not a single cointegration test. No Granger test. :((
For example, let's look at hypothesis testing... http://alglib.sources.ru/hypothesistesting/
Not a single unit root test. Not a single cointegration test. No Granger test. :((
You say a lot of right things about R, but you see its advantage over ALGLIB/FANN in the wrong way.
The main advantage of R as a language and development environment is that it allows for enormous productivity in prototyping trading systems. For that time that amateurs of programming spend on binding a couple dozens of ALGLIB methods, you can perform several dozens of experiments in R (equity calculation based on buy/sell signals with allowance for spreads and commissions is made literally in five lines).
In addition, to speed up slow calculations, there are very simple and easy to use parallel programming tools.
R is much slower than C++/MQL5, so the final version of the strategy should be written in something else.
Porting R or its libraries to MQL5 makes no sense, but the MQL5-R interface could be useful. If I have to choose what to combine it with (R/MATLAB/MATHCAD), I would definitely go for R.
The main advantage of R as a language and development environment is its enormous productivity in prototyping trading systems
The main advantage is a specialized package, and what you write is the result of this specialization.
When using specialised packages, it's very important to see "what happens". For example, compare Statistics and EViews. The latter is an encyclopaedia of "what happens", while the former is a set of tools. We compare EViews and Matlab. Matlab compared to EViews is also an encyclopaedia, but only a very competent person would be able to see that, for sure EViews was useful and STATISTICS is useless.
Of course, an econometrician, who has eaten not only the tail of the dog but also the dog itself, will use a paid (?) Matlab. A qualified person is unlikely to rewrite one library from one language to another without good reason.
And again: the availability of a large volume of books on statistics, BP, econometrics in relation to R and R code.
By the way, the ideology of using R is very similar to that used by Metacquotes: free tools and a huge free codebase with free articles. Only in R it is much wider.