From theory to practice - page 137

 
Alexander_K2:
That's not what he meant. If you look closely at the graph

you will see that in the second graph the variance changes slightly as the window shifts. This is the manifestation of the non-stationarity of the process, whose tails we need to pick up and capitalise on.

If, on the other hand, we consider the average parameters of this window when the sample size is gigantic, then they remain virtually unchanged.

Actually, there's something wrong with this picture. STO can't go that way.
 
Yuriy Asaulenko:
Actually, there's something wrong with this picture. STO can't go like that.
That's a question for Koldun. When I saw it, I nearly fell off my chair myself, but my father-in-law held me back and literally forced me to get on with my work, rather than get into mysticism or drunkenness. My job is to record the fact
 
Alexander_K2:
He meant a little differently. If you look closely at the graph

you will see that in the second graph the variance changes slightly as the window shifts. This is the manifestation of the non-stationarity of the process, whose tails we need to pick up and capitalise on.

If we look at the average parameters of that window when the sample size is gigantic, they barely change at all.


Hmm, you have the sampling window depending on the sampling variance, and as far as I understand the variance depends on the sampling window. Or by "sampling variance" do you count the variance on the whole available history?

PS Sheet1.

 
Nikolay Demko:

Hmm, your sample window depends on the variance of the sample, and as far as I understand the variance depends on the sample window. Or by "sampling variance" you mean the variance on all available history?

On all the history available to me.

 
Alexander_K2:
That's a question for Wizard. I almost fell off my chair when I saw it, but my father-in-law held me back and literally forced me to get on with my work, rather than get into mysticism or drunkenness. My job is to record the fact.
What's there to fall back on? the graph on which the STO counts is in the picture. The STO period is small, as you can see. Where it should be decreasing, it somehow increases, for no reason at all.
 
Alexander_K2:

All over.


Do you realise that this is a peek into the future?

In history this kind of thing works, in real time it doesn't.

 
Alexander_K2:
That's a question for Wizard. I almost fell off my chair when I saw it, but my father-in-law held me back and literally forced me to get on with my work, rather than get into mysticism or drunkenness. My job is to record the fact


 
Nikolay Demko:

Do you realise that this is a peek into the future?

In history such tricks work, but not in real time.

I think it's time to get with the program, Nikolai. Or do you still have doubts?
 
Nikolay Demko:

Do you realise that this is a peek into the future?

This kind of thing works in history, but not in real time.


I can count the waving, move it back half a period, and it will be smooth and not laggy in the story.

This is an example of looking into the future.

ZZY I won't have values for the last half period before realtime, but that's such nonsense for graal builders that it's not even worth bothering with ))

ZZZY For the first point of the sliding window the data to the right is not yet available, but it is already accounted for in the waving, standard deviation and variance.

 
Alexander_K2:
I think it's time to get with the program, Nikolai. Or is there anything else you doubt?

While there is no deep ticking history, and there is a glimpse into the future in the calculation, I am very much in doubt.