From theory to practice - page 623

 
secret:
In 600 pages the commander still hasn't grasped that going back to the mean is in no way dependent on the form of the increments.)

)))
He does not understand what a graph of distributions means.

 
Novaja:

In general I see, the kurtosis is reduced, tails are picked up ---> from Laplace to normal, from normal to uniform. Then what's at the beginning? Not Laplace? What? It is easily described by an exponent, if we take one side. These are the EURUSD window-monthly minutes.


I'd like to see the stats on the minutes... What's the excess there?

In the beginning, when I was working with pure ticks, there was no excess~3 and asymmetry~0. Ticks - to hell with it!

 
Alexander_K2:

I'd like to see the statistics on the minutes... What's the excess there?

In the beginning, when I was working with pure ticks - there was no kurtosis~3 and asymmetry~0. Tics, screw it!

On the minutes.


Column 1
Average -0,011039558417663
Standard deviation 0,05793349275103
Moda 0
Median 0
First quartile -6
Third quartile 6
Dispersion 200,655772679815
Standard deviation 14,1653017151
Skewness 22,6682647496561
Skewness -0,007829129190145
Range 620
Minimum -375
Maximum 245
Amount -660
Amount 59785

This is the CLOSE increments of the EURUSD (two months), converted into pips.

 
Novaja:

In minutes


Column 1
Average -0,011039558417663
Standard deviation 0,05793349275103
Moda 0
Median 0
First quartile -6
Third quartile 6
Dispersion 200,655772679815
Standard deviation 14,1653017151
Skewness 22,6682647496561
Skewness -0,007829129190145
Range 620
Minimum -375
Maximum 245
Amount -660
Amount 59785

These are the CLOSE increments of the EURUSD minutes, converted to pips.

Pips. No options. Only Erlang flows can comfort me.

 
Alexander_K2:

I'd like to see the statistics on the minutes... What's the excess there?

In the beginning, when I was working with pure tics - there was no excess~3 and asymmetry~0. Tics - fuck it!

well thank god! finally ;)))

 
Alexander_K2:

Screw it. No choice. Only Erlang streams can comfort me.

running in circles... :(

I thought you'd gotten rid of it... but no...

 
Alexander_K2:

Screw it. No choice. Only Erlang streams can comfort me.

So what's next? If the origin is unknown and what it affects, why isn't the law of large numbers enforced? There are more questions than answers.

 

Here are two more 60,000 minute files for research.

1 file of the increment of the transformed V.R.

2 file of incremental difference between the current and transformed series.

 
Evgeniy Chumakov:

Here are two more 60,000 minute files for research.

1 file of incremental transformed V.R.



Column 1
Average -0,009283333333333
Standard deviation 0,164330498707676
Moda 0
Median 0
First quartile 0
Third quartile 0
Dispersion 1620,27076833081
Standard deviation 40,2525871010896
Skewness 18,8175273500745
Asymmetry 0,007821686703684
Range 540
Minimum -291
Maximum 249
Amount -557
Amount 60000
 
Evgeniy Chumakov:

Here are two more 60,000 minute files for research.


2 file the incremental difference between the current row and the converted row.


Column 1
Average -0,00095
Standard deviation 0,154547684159877
Moda 0
Median 0
First quartile -6
Third quartile 6
Dispersion 1433,09920075088
Standard deviation 37,8562967120514
Skewness 16,1201747750326
Skewness -0,001973043862113
Range 446
Minimum -217
Maximum 229
Amount -57
Amount 60000