Machine learning in trading: theory, models, practice and algo-trading - page 209
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
...
Note that none of the fervent fans of R are not happy about the appearance of analogous functions in MQL, they don't need it now and will never need it, no matter how great the library in MQL is. Because these people are fanatics in the worst sense of the word, because they don't care whether they think R libraries are right or not, because someone with academic degrees said that they think right, and that's enough to completely trust the authority of these scientists, enough to stop seeing through their noses, enough to unconditionally believe the black box and think that if it can not work in R, it can not work anywhere else.... This is Religion, eR-brain....
On my part, I express my deep gratitude for providing such a great opportunity to stat and matte calculations, for the ability to easily and quickly test ideas and implement them in your robots, thank you for the meticulous and methodical study of you and your team methods, in an effort to provide users with the highest quality and speed even in spite of the recognized authorities in the field of machine learning and stat and matte calculations. A wish only small, please do not pursue the goal of a complete copying of R, fans of R do not need it for the above mentioned reasons, and the rest certainly do not need it. I want to see full-fledged instruments, accurate and fast and without reference to universally recognized authorities. History tells us that this is the only way to move to the next stage of development, and history tells us that MetaQuotes has always gone this way, which allowed it to achieve outstanding results in its trading platform. Copying can only give the same result, but not better. If you accept the arguments of R-astes, it means to remain forever at their level, in their close and dark world of delusions....
Note that none of the fervent fans of R are not happy about the appearance of analogous functions in MQL, they do not need it now and will never need it, no matter how great the library in MQL is. Because these people are fanatics in the worst sense of the word, because they don't care whether they think R libraries are right or not, because someone with academic degrees said that they think right, and that's enough to completely trust the authority of these men of science, enough to stop seeing further than your nose, enough to unconditionally believe the black box and think that if it can't work in R, it can't work anywhere else.... This is Religion, eR-brain....
On my part, I express my gratitude for providing such a great opportunity to stat and matte calculations, for the ability to quickly and easily test the ideas and implement them in your robots, thank you for the meticulous and methodical study of you and your team methods in an effort to provide users with the highest quality and speed even in spite of the recognized authorities in the field of machine learning and stat and matte calculations. A wish only small, please do not pursue the goal of a complete copying of R, fans of R do not need it for the above mentioned reasons, and the rest certainly do not need it. I want to see full-fledged instruments, accurate and fast and without reference to universally recognized authorities. History tells us that this is the only way to move to the next stage of development, and history says that MetaQuotes has always gone this way, which allowed it to achieve outstanding results in its trading platform. Copying can only give the same result, but not better. If you accept the arguments of R-astes, it means to remain forever at their level, in their close and dark world of delusions....
http://www.wolframalpha.com/input/?i=integrate[pdf[gammadistribution[0.5,1],x]+,{x,0,1*10^(-90)}]
The integral is the area of the figure shaded in blue. As you see, the left side of the shaded figure tends to infinity. Even though wolfram does not include the point x=0 in the pdf function, there is still no finite "highest point," you can think of the left side of the figure as growing up infinitely. Logically, if the left side of the figure grows up infinitely, then its area will also tend to infinity. But this doesn't actually prevent you from getting a non-infinite result when determining the area of that figure. Math.
The problem is that Wolfram Alpha (and Matlab) "hand-defines" density at zero as 0, i.e. they consider the expression valid only for x>0.
This allows you to determine the density at all points and get rid of the uncertainty in determining the density.
For this reason, by integrating in Mathematica and Matlab you don't get problems.
Try doing the same by integrating in R, i.e. using dgamma. What would the cdf(x) calculated this way look like?
Try doing the same thing by integrating in Excel a function that returns nan at zero. What would the cdf(x) calculated this way look like?The problem is that Wolfram Alpha (and Matlab) "hand-defines" density at zero as 0, i.e. they consider the expression valid only for x>0.
This allows you to determine the density at all points and get rid of the uncertainty in the definition of density.
For this reason, integrating in Mathematica and Matlab you don't get problems.
Try doing the same by integrating in R, i.e. using dgamma. What would the cdf(x) calculated this way look like?
Try doing the same thing by integrating in Excel a function that returned nan at zero. What would the cdf(x) calculated this way look like?Reading your and Renat's posts, I'm just stunned.
Several people on the forum are saying: "How nice that in MKL5 they port statistics from the world's top system R".
And we are countered with, "No we only have it partially referring to the top!"
Why would you do that?
Ported, brought proof that all properly ported. After that it is clear that this part of ICL corresponds to the world's top in the field of statistics, which is R, and which has not been Malab and has never been the Math package for five years.
Did you find something? No problem. Write to R. If you convince them, then make the edit following R and start PRing your level, which has been recognized by an authority like R.
If R stays with his opinion, you keep your mouth shut.
Now you are trying in every possible way to question that your statfunctions are relevant to the world top.
That's the point of all your posts.
And the fact that two people on the forum started to prove to you something there in infinity, I have translated to you the meaning of how it is perceived by the majority: "Not quite R".
Try doing the same thing by integrating in R, i.e., using dgamma. What will the cdf(x) calculated this way look like?
In R, next to that shaded figure will be another rectangle. Width 0, height infinity (with coordinates [0,0]->[0,Inf], I would even call it an infinite line). To the result of the integration from that link we need to add the area of such a rectangle, this would be the visualization of the result of R.
What is the area of a rectangle with width 0? What is the area of a line? The questions are very strange, and the answer is clearly not infinity. The result 0*Inf is undefined, but in this case I think the answer is 0.
In Excel it will be a rectangle with width 0, height NAN. I don't know what its area will be :D
In the optimization algorithm contest, you had a great chance to show
I didn't really follow that contest, I remember only a hundred pages of organizer's rubbish, changing the rules on the fly, and other fun stuff. Who won there though? Winner seemed to be even promised a cash prize from MQ, who was the lucky one?
I did not really follow that competition, I remember only a hundred pages of flub from the organizer, with changes in the rules on the fly, and other joys. Who won there though? The winner seemed to be even promised a cash prize from MQ, who was the lucky one?
San Sanych, how you try to belittle everything.
You don't even hesitate to call Matlab, Wolfram, and Mathematics "I don't know who that is". All you want to do is throw in and go,"Your statfunctions are NOT relevant to the world top."
They all do and we have proven it by including the source code of unit tests of correctness, accuracy and benchmarks. And we have shown that our implementation in MQL5 is many times faster than C++ implementation of the same functions in R.
And we have shown errors in R as well. Or have you already forgotten how you quietly accepted the first error on the accuracy of calculations and the use of a weak calculation function?
In R next to that shaded figure will be another rectangle. Width 0, height infinity (with coordinates [0,0]->[0,Inf], I would even call it an infinite line). To the result of the integration from that link we need to add the area of such a rectangle, this would be the visualization of the result of R.
What is the area of a rectangle with width 0? What is the area of a line? The questions are very strange, and the answer is clearly not infinity. The result 0*Inf is undefined, but in this case I think the answer would be = 0.
In Excel, it will be a rectangle with width 0, height NAN. I don't know what its area will be :D
San Sanych, how you try to belittle everything.
You don't even hesitate to call Matlab, Wolfram, and Mathematics "I don't know who that is". All you want to do is throw in and go,"Your statfunctions are NOT relevant to the world top."
They all do and we proved it by including the source code of unit tests of correctness, accuracy and benchmarks. And we have shown that our implementation in MQL5 is many times faster than C++ implementation of the same functions in R.
And we have shown errors in R as well. Or have you already forgotten how you quietly accepted the first error regarding the accuracy of calculations and the use of a weak calculation function?
Yes, you know better.
Understand. Or you belong to the world top, and this you prove by your comparisons, or you do not belong to it, and you yourself point out the differences between your implementation and the world top. Your not wanting to be in the world top, being in it, just amazes me.
PS.
Do not even dare to call Matlab, Wolfram and Mathematics "I do not know who it is.
Give me a link to rankings of statistical packages that had Mathlab (Wolfram) in them. Matlab was, but has died. I gave in my blog on your site and many times laid out on the forum
Yeah, you tell me.
Understand. Either you belong to the world top, and this you prove with your comparisons, or you do not belong to it, and you yourself point out the differences between your realization and the world top. Your not wanting to be in the world top, being in it, just amazes me.
PS.
Do not even dare to call Matlab, Wolfram and Mathematics "I do not know who it is.
Give me a link to rankings of statistical packages that had Mathlab (Wolfram) in them. Matlab was, but has died. Here I give in my blog on your site and many times posted on the forum