From theory to practice - page 1163

 
Unicornis:

Well forgive the villeins, they don't know that you're prettier than the word 'daddy' churned out of you by xforex with rhythmic bruzers-style moves. ;)

What a child of feeble-mindedness :))) Your posts should be plastered over toilets.

 
Alexander_K:

The idea is that this non-linearity of time between data is taken into account in my variance calculations. If there is no data, there is obviously an error. I can report that, oddly enough, the market is quite accurate. And mistakes are expensive.

I get it, I mean the timing of the next tick is randomly generated, i.e. 2 identical systems will show different results

Or is that not critical there?

 
Maxim Dmitrievsky:

I get it, I mean the timing of the next tick is randomly generated, i.e. 2 identical systems will show different results

or is that not critical there?

On different inputs, yes. I am striving for the TS at any broker to work the same way. This requires synchronization of flows between different DCs.

I take the ticks as follows:

1. Every N seconds I get a quote.

2. I look if Bid, Ask or server time of this quote has changed.

If any of these parameters has changed - the quote is real (i.e. we have an event) and it is used in calculations, if not - it isn't.

At the end of the week I compare my set of events with quotes from Dukas.

Now I have desynchronisation due to strange dips at certain frequencies.

If there is inaccuracy in data reception/processing, then the error in process variance calculation is obvious.

 
Alexander_K:

I don't want to suspect DC of playing games with quotes and their flow, but taking this into account and making adjustments is simply necessary.

Alexander_K:

From the child of the feeble mind :)).

DC does whatever it wants with the quotes flow and whenever it wants, and it doesn't have to report it to anyone, including you. Look in the mirror and smile more often. ;)

 
Unicornis:

The brokerage house does whatever it wants with the quotes flow and whenever it wants, but it does not have to report it to anyone, including you. Look in the mirror and smile more often, and you will see your grail truth. ;)

:)) I am not complaining, write what you like. However, being a nephew of podotr (by the style of presentation), you could write something smarter and not dishonor your uncle.

We're looking for the Grail, not a toilet. Aren't we?

 
Alexander_K:

Oleksiy cannot do anything with his hands, or even with his head. He only consumes alcohol with it, which is obvious to everyone.

Maybe he's found his truth.

 
Alexander_K:

Oleksiy cannot do anything with his hands, or even with his head. He only consumes alcohol with it, which is obvious to everyone.

It's obvious to me that your rake continues)

 
If you need the Grail, you cannot do without taking into account the changing times of market moves. The time directly affects the accuracy of trades.
The timing of the defining market movements is constantly changing.
Accordingly, the nature of the market changes and so do the ways in which we profit from those movements.
Yes, Sasha is right about time non-linearity. Only he didn't make the main point:)
Although you are heading in the right direction.
Translate the non-linearity of motion into a non-linearity in time, perform a two-dimensional filtering and you'll get what you're looking for.
 
Alexander_K:

On different inputs, yes. I'm aiming for the TS at any broker to work the same way. This requires synchronisation of threads between the different DCs.

I take the ticks as follows:

1. Every N seconds I get a quote.

2. I look if Bid, Ask or server time of this quote has changed.

If any of these parameters has changed - the quote is real (i.e. we have an event) and it is used in calculations, if not - it isn't.

At the end of the week I compare my set of events with quotes from Dukas.

Now I have desynchronisation due to strange dips at certain frequencies.

If there is inaccuracy in data reception/processing, the error in process variance calculation is obvious.

Well, then all the quotes will be recognized by you as real, if only because their server time differs by N seconds.

No quantization - no grail. On your front, it's different for me.

Read the classics - why didn't Kotelnikov please you?

 
Martin Cheguevara:
If you need the Grail, you can't do without taking into account the time changes determining the market movements. The time directly affects the accuracy of transactions.
The timing of the defining market movements is constantly changing.
Accordingly, the nature of the market changes and so do the ways in which we profit from those movements.
Yes, Sasha is right about nonlinear time. Only he didn't make the main point:)
Although I was going in the right direction.
Translate the non-linearity of motion into a non-linearity in time, perform a two-dimensional filtering and you'll get what you're looking for.

If I understand correctly, we are talking about the distribution of events ( the size of a candle or a tick) on the x-axis (time).