From theory to practice - page 1426

 
Yuriy Asaulenko:

Why collect and analyse seconds or ticks? - the above showed that minutes are enough.
We want to enter the trade better - we start analysing those very ticks, when the minutes show that it is time to enter the trade. Before that, we don't care and forget about it.
Otherwise, it turns out - from the cannon to the sparrow.
I already gave you an example, it's like estimating the trajectory of an aircraft by measuring the vibrations of its fuselage.

I don't work specifically with ticks, but take them as a basis.

Then I look at time intervals between ticks (Doc showed a picture in this thread - there's a gamma distribution), then I start to thin them, i.e. drop every 2nd, 3rd, etc. I see different pictures. For example:

With any type of thinning, equal time intervals between quotes are never obtained!

I.e. OPEN/CLOSE M1, M5,... - This is an incorrect discretization of the process that contradicts the very nature of the market.

Therefore, I work with time intervals, which I set artificially with Erlang distribution. This is more suitable for real data.

I'm strongly against working with OPEN/CLOSE M1, M5,... although I won't insist on my particular method of data reception.

 
Alexander_K:

I don't work specifically with ticks, but take them as a basis.

Then I look at time intervals between ticks (Doc showed a picture in this thread - there is a gamma distribution), then I start to thin them, i.e. drop every 2nd, 3rd, etc. I see different pictures. For example:

With any type of thinning, equal time intervals between quotes are never obtained!

I.e. OPEN/CLOSE M1, M5,... - This is an incorrect discretization of the process that contradicts the very nature of the market.

Therefore, I work with time intervals, which I set artificially with Erlang distribution. This is more suitable for real data.

I'm strongly against working with OPEN/CLOSE M1, M5,... although I won't insist on my particular method of data reception.

Sash, here's the H4.

It's OK.

Just finished:


 
Renat Akhtyamov:

Sash, here, H4.

It's OK.

Just finished:


Nah, well, if it works out with uniform sampling, I won't change my mind, of course.

The question of the method of taking quotes is a very debatable one.

 
Are you going to bed? ;)
 
Evgeniy Chumakov:
Are you going to bed? ;)

No, I'm just finishing up, because the signal's already been pissed off.

And I say the FormulaE is a thing of the past.

;)

 
Alexander_K:

No, well, if it works with uniform sampling, I won't change my mind, of course.

The question of the method of receiving quotes is a very debatable one.

I don't accept them, they themselves are sort of already in the terminal, automasyski

;)

 
Evgeniy Chumakov:
Are you going to bed? ;)

No way.

No time. I need the Grail at all costs! I need +50% a month at all costs.

 
Alexander_K:

I.e. working with OPEN/CLOSE M1, M5,... - is an incorrect process discretisation, it contradicts the nature of the market itself.

The fact that you dream of seeing tick arrivals in a "photon field" doesn't change the principles of time sampling - on the horizontal axis is time, on the vertical axis is a value of the observed parameter - nothing more or less.

Alexander_K:

I am strongly against working with OPEN/CLOSE M1, M5,... although I will not insist on my method of data reception.

don't use it, why shout about it? ... You're smoking something strong... ask your loved ones to keep you away from hallucinogens, household chemicals, etc... - I saw a famous person on TV who used a clogging agent because it was freely available...

 
Igor Makanu:


don't use, why shout about it? ... ...ask your loved ones to keep you away from hallucinogens, household chemicals, etc... -


I think if a doctor looks at this thread, he'll conclude that everyone is consuming.

 
Evgeniy Chumakov:

I think if a doctor looks at this thread, he will conclude that everyone is consumed.

Do what you want, I'm going to bed.