From theory to practice - page 548
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Alexander, I decided to work with your old data here:https://www.mql5.com/ru/forum/221552/page296#comment_6994035
The question there was whether the exponent fit into your time intervals, there were also increments, do you remember how you did them? Were they uniform increments or exponential? That's important. It was logarithmic on the time intervals, the exponent didn't fit, and it didn't fit on the increments either, the exponent didn't fit there.
The theory is good, but the practice fails)
Alexander, I decided to work with your old data here:https://www.mql5.com/ru/forum/221552/page296#comment_6994035
The question there was whether the exponent fit into your time intervals, there were also increments, do you remember how you did them? Were they uniform increments or exponential? That's important. You got the logarithmic on the time intervals, the exponent didn't fit, the same goes for the increments, the exponent doesn't fit there.
Alexander, I decided to work with your old data here:https://www.mql5.com/ru/forum/221552/page296#comment_6994035
The question there was whether the exponent fit into your time intervals, there were also increments, do you remember how you did them? Were they uniform increments or exponential? That's important.
There the data was collected every 1 second as I recall.
By the way, and Doc (wasn't it with you??) was doing research on time intervals between tick quotes. There's something like "dirty" Erlang flow of 2nd order (Doc even derived a formula - Gamma+Koshi. Where is it now? Probably earning his deposit by hard work too...). All in all, a non-Markovian process.
That's why I'm forced to work on an exponential scale. I set it with a geometrically distributed generator. This is the only way to get to Laplace in increments - and nothing else.
Alexander, I decided to work with your old data here:https://www.mql5.com/ru/forum/221552/page296#comment_6994035
The question there was whether the exponent fit into your time intervals, there were also increments, do you remember how you did them? Were they uniform increments or exponential? That's important. You got the logarithmic on the time intervals, the exponent didn't fit, the same goes for the increments, the exponent doesn't fit there.
There the data was collected every 1 second, as far as I remember.
By the way, Doc (not together with you??) also did research on time intervals between tick quotes. There's something like a 2nd order Erlang "dirty" flow (Doc even derived a formula - Gamma+Koshi. Where is it now? Probably earning his deposit by hard work too...). All in all, a non-Markovian process.
That's why I'm forced to work on an exponential scale. I set it with a geometrically distributed MF generator. This is the only way to get to Laplace in increments - and nothing else.
I.e. uniform intervals, 1 sec, not exponential? Throw me or here the data where you take exponentially, intervals and returns
If you come up with any conceptual ideas, post them, OK?
I'm busy now - but I read the forum. I'm waiting for someone from the Hilbert space (like Aleshenka, Vladimir or Koldun) to write something congenial.
I.e. uniform intervals, 1 sec, not exponential? Throw me the data where you take exponentially, intervals and returns
There, if returns and time =0, it's a pseudo-quote, otherwise it's real.
I'll throw it on Saturday-Sunday.
If you come up with a conceptual idea, post it, OK?
I'm busy now - but I read the forum. I am waiting for someone from Hilbert space (like Aleshenka, Vladimir or Koldun) to write something congenial.
That's the thing, I need your data with an exponential reading. I'll search the forum myself.
That's just it appeared, I need your data with an exponential reading. I'll search the forum myself.
Here is the Laplace, in blue, the blue-returns Close distribution for 2 months, 57 thousand data, the red-blue exponent is almost perfectly fit, except for the "tails". You don't have it. It's on a logarithmic scale, I love it, it's much clearer to see.
I think I have the same, maybe a little better.
However, so what? We have laplace motion, which unlike the Wiener process, has been little researched.
If we apply the mathematics of the Wiener process, we have a net +0% gain.
We need a conceptual breakthrough.
It's like: "Uncles! Did you know that..." followed by a little genius text.