From theory to practice - page 1698

 
Дмитрий:
Well, of course! Calculating the distribution function, variance and other statistical process parameters is alchemy.... Well, or chiromancy!

It is, if the sample is not taken for a sequence of independent equally distributed variables (Glivenko-Cantelli theorem)

 
Aleksey Nikolayev:

That is, if the sample is not taken for a sequence of independent equally distributed variables (Glivenko-Cantelli theorem)

The practical implementation of the theorem is based on hypotheses.

Only mathematics is based on explicit proofs, and not all of them.

 
Aleksey Nikolayev:

Exactly, if the sample is not taken for a sequence of independent equally distributed variables (Glivenko-Cantelli theorem)

I read on this topic yesterday on Habra : "Intuitive explanation of hypothesis testing and the p-value"https://habr.com/ru/post/475048/

 
Дмитрий:

The practical implementation of the theorist is based on hypotheses.

Only mathematics is based on explicit proofs, and not all of it.

We are talking about iid tests

 
Vladimir Baskakov:
Do you really think it is a channel? A channel in forex is something horizontal, parallel

The channel is an important, indispensable thing in analysis. I don't think there are any more impressive ways of determining in the direction of market interest.

A channel is a market set-up. It can change angle, it can change direction depending on the mood of the market. But it will always exist.

And a horizontal channel is a flat. )) You'd better ask Sasha. He always has a horizontal channel.

 
Uladzimir Izerski:

The channel is an important, indispensable thing in analysis. I don't think there are any more impressive ways of determining in the direction of market interest.

A channel is a market set-up. It can change angle, it can change direction depending on the mood of the market. But it will always exist.

And a horizontal channel is a flat. )) You'd better ask Sasha. He always has a horizontal channel.

It's like K. Marx taught you.
 
Vladimir Baskakov:
It's like K. Marx was your teacher.

I barely have the patience to answer your question.)

And you don't have the basic intelligence to hear the answer.

Bye-bye, you little punk.)

 
Uladzimir Izerski:

I barely have the patience to answer your question.)

And you don't have the basic intelligence to hear the answer.

Bye-bye, you little punk.)

I like reading your posts, I like the naiveté.
 
Maxim Dmitrievsky:

Now that's another conversation. Gave you the gift of free dopamine, too. Enjoy it.

I'll post the code in a little while. You can't now.

For now it is enough to see the trend or possible start of a flat.

I haven't seen any ready-made solution for predicting a flat.

 
Igor Makanu:

I read on this topic yesterday on Habra : "Intuitive explanation of hypothesis testing and the p-value"https://habr.com/ru/post/475048/

Usually such a problem is solved with Fisher's exact test. The statistics of this test are distributed according to a hypergeometric law. Calculating the CDF (to find the p-value) of this distribution in the MT5 statistical library produces a division error of zero.

Reason: