The fight against PYTHIA 8 has come down to a dumb formula...

 

ABS(((High-Low)/(Open-Close))*(Vol*(Open-Close))/1000...

the meaning is not clear yet, but practically every following period works in 80% of cases...:-))

(divided by 1000 to get digits that are digestible and can be immediately covered by a glance)...

What is it...either the inertia of a spontaneous trend or something else...somehow everything is dumb...expected more...

 

This is this programme:

PYTHIA — программа моделирования процессов столкновения элементарных частиц при высоких энергиях на ускорителях элементарных частиц методом Монте-Карло. Программа разработана группой физиков Лундского университета в Швеции.

Development of PYTHIA began in 1982. In 1997 the project merged with another related development, JETSET, to combine their functionality into one generator. Versions of the program up to version 6 were implemented in Fortran 77. However, since 2004, PYTHIA 8 began to be developed in C++. As of January 2010, both are being developed in parallel.

...

(http://home.thep.lu.se/~torbjorn/Pythia.html)

At least tell us what you have faced.

 
zoritch:

ABS(((High-Low)/(Open-Close))*(Vol*(Open-Close))/1000...

and this too can be reduced to the even simpler formula ABS(High-Low)*Vol *SIGN(Open-Close)

so what of this?

 
Farnsworth:

This is this programme:

(http://home.thep.lu.se/~torbjorn/Pythia.html)

At least explain what and what you were running into.


I've already said that...:-))) The main thing is that if the condition set to the previous two...er...candle...periods, works stupidly for the bigger-lower for the current candle...

the direction is not important...so it seems to be working...i even ran it timidly on the real...so far everything seems to be OK...:-))

nothing collided here, of course, and built an imaginary model for the entire period (ahead of the real one....nu I took from 2000) and compared it to the real one...

(closer probably to brown noise...but that's not so important...a dozen billion I ran...)

 
zoritch:


if the condition set to the previous two candlesticks works stupidly on more-less for the current candlestick...

direction doesn't matter

an imaginary model for the whole period is built (ahead of the real one.... by 2000) and compared to the real one...


does anyone understand what is written here ?
 
if these values compared to the two previous bars, candlesticks, whatever your terminology is, perform a blunt more-or-less, then the direction of the future candlestick is quite predictable...
 
sergeev:

and this too can be reduced to the even simpler formula ABS(High-Low)*Vol *SIGN(Open-Close)

and what of it?


I'm not talking about maths...of course...that's not what I meant...the formula was not made by me, but by a machine...:-))
 

Nothing to say about my karma...just sharing the joy of seeing some sort of pattern...

after R:Base (which flew the first shuttles), in a fortnight produced the first result...:-))

(P.S. frantically large volumes were missed, as it is useless to understand the psychology and motivation of large players...

i want to note that it didn't depend on the timeframe either...)

 
Farnsworth:

This is this programme:

(http://home.thep.lu.se/~torbjorn/Pythia.html)

At least explain what and what you were running into.

I have eight bolted on R:Base...a relational software competing with Oracle at one time....

don't see any problems... the system picks patterns out of random signals, then just dumbly processes them in the RDBMS...

 

how to delete the whole topic...fuck it...:-)))

(don't, so don't... better wave analysis... especially on Friday...)

 
zoritch:

I have eight bolted on R:Base...a relational software that competed with Oracle at one time....

I don't see a problem... the system picks patterns out of random signals, then just dumbly processes them in the RDBMS...


And in general you are at least approximately familiar with the dynamics of formation of the ultrarelativistic proton... I think the laws of shrinkage (God forgive me) of the gluon cloud

similar to the dynamics of absolutely any process... there's no time there... and here it can, in principle, be thrown away...

Reason: