Trying to distinguish between real currency mark-up and illusory currency mark-up and recognising the simultaneity of the pairs' moves. - page 5
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
https://www.youtube.com/watch?v=s2H3UG7Nab4 video link, can't get it in.
♪ if you don't mind, put it in here ♪
I have to go out for a while, I'll be back later to discuss the video...
>
trol222:
might come in handy... have a look...
Renat 24.08.2006 09:08
The simplest and rudest way for automatic calculation of tick volume (for example for M1):
volume = (Open - Low + High - Low + High - Close) * pow(10,Digits) + 1
https://www.mql5.com/ru/forum/54216
About ticks and pipsing...... I think that ticks are not necessary for pipsing....
Just imagine a balloon with many molecules inside it. Some molecules fly forward and hit the walls of the ball. And if pressure of these impacts is more to the right - ball moves to the right, if to the left - to the left.... It is clear that at first sight such small movements are quite hoaty.... So, if we analyse the moments of such molecules (when they hit the walls) and their duration, we can make assumptions about the nature of the movement of the ball (whether it continues or stops)
.
Anyway, the balloon analogy is very good, from statthermodynamics. Here's looking at the system statistical-thermodynamically. Tick volumes... Of course, they aren't bad, but they won't tell you anything important about the movement: a large tick volume may be both trending and flat. It is better to look at the increase of rates for a certain time (say, for the timeframe), it will be more informative.
Now - briefly in essence. Code the movements of pairs by one of three classes:
-1 - strong downward movement,
0 - weak movement,
+1 - strong upward movement.
Boundaries between moves are best calculated from the quantiles of the probability distribution of returns, breaking it down into three quantiles. Or just from nothing if you are too lazy.
Take into account the position of the analyzed currency pair in order to put it in the correct class. For example, if you analyse the franc, and the USDCHF pair is rising strongly (+1), the franc will fall, i.e. it will be -1 for the franc.
Next, make a vector of the state of the gas. Suppose we have nine chif pairs, we get, say, the following numbers: <+1,0,-1,0,+1,-1,0,0,+1>. This is the microstate of the gas, with the class of a particular pair at each position. All possible microstates of a gas are equally probable.
The macrostate of a gas is a set of equivalent microstates. For the macrostate of the gas in this sequence, the order of these numbers is actually unimportant. Only the quantities of different classes are important. This is where you get different probabilities. There's three +1, two -1 and four 0. Something very similar to a chiff flatt, by the way. That's what it is. A flat is usually the most likely macro state.
The thermodynamic probability of that macro-state (the number of equivalent micro-states - see underlined) is calculated through factorials: 9!/(2!*3!*4!)=1260. (By the way, the most probable macrostate is the perfect chif flatt: 9!/(3!*3!*3!)=1680. The perfect flat in this case is three +1s, three 0s and three -1s).
And, say, a microstate <+1,+1,0,+1,+1,+1,+1,+1,0,+1> corresponds to a macrostate with thermodynamic probability 9!/(7!*2!)=36. And this macro-state, as we see, is 35 times less likely than the previous macro-state. It is actually very similar to a trend.
A trend is a rare macrostate. The trend on a pair is an interesting dick, but it is only reliably registerable when multicurrency is taken into account. A trend can only be reliably registered by currency. A flat is trickier as it is more complex. Actually, there are several types of flats, but I will not go into details.
By the way, if you take the logarithm of thermodynamic probability, you get almost entropy. This entropy will probably follow me all my life: then informational (in a branch about feature selection), then physical :).
You can do your own digging from here. If you have any objections or interesting ideas, please write to me. Only, please, do not ask idiotic questions about what is quantile, how factorials are calculated or why probability is counted through factorials. You don't look like someone who needs to bring everything on a silver platter. Try reading this nonsense a few times and make sense of it. It took me a few months to crystallise this bullshit in my brain.
P.S. We could go into some very rambling about phase transitions (gaseous/liquid/dead flats and crystalline trends), but that's enough for the first statement. And of course there are plenty of pitfalls too.
Damn, wrote a thorough reply - and it's all gone when I hit the "Add comment" bar. It's because the session time is too short.
Anyway, the balloon analogy is very good, from statothermodynamics. Here's looking at the system statistical-thermodynamically. Tick volumes... Of course, they aren't bad, but they won't tell you anything important about the movement: a large tick volume may be both trending and flat. It is better to look at the increase of rates for a certain time (say, for the timeframe), it will be more informative.
Now - briefly in essence. Code the movements of pairs by one of three classes:
-1 - strong downward movement,
0 - weak movement,
+1 - strong upwards.
1- it is clear that tick volume can be high both in a trend and in a flat, but you must have misunderstood me.... 2- If we shift to ticks I emphasize that the frequency of tick arrivals (volume) and price changes cannot be considered separately - they must be combined.....
2 - I don't think the way you suggest would be much different from the commonly available Semenych indicators, in my example with the ball (I wrote it first with the junkie, at least I got the typo))) I was drawing attention to the need to separate the molecules that hit the ball from the total mass (I have not yet figured out how).
Now I'm interested in velocities and fans... and also the transition sections from ff to ff... let's see
Valery, try to answer in such a way that it is clear what is quoted and where the answer to the quote is. It's clear to me (today), but not to the rest of us.
trol222: 2- предложенное вами сваливание в одну кучу думаю врядли будет отличатся особо от общедоступных индикаторов семеныча, в том моем примере с шариком(написал сначала с нариком, ладно хоть опечатку увидел))) я делал упор именно на то что нужно как то отделить от общей массы те молекулы которые именно ударяют о стенки шара (как не придумал еще) .
It is different, and cardinally.
Notice the difference?
I wanted to write about Semenych from the start, but changed my mind at the last moment... Okay, that's it. First, try to manage to write your own thoughts.
Damn, wrote a thorough reply - and it's all gone when I hit the "Add comment" bar. It's because of too little session time.
Damn, I wrote a thorough reply - and it all disappeared when I hit the "Add comment" bar. It's because of too little session time.
Anyway, the balloon analogy is very good, from stattermodynamics. Here's looking at the system statistical-thermodynamically. Tick volumes... Of course, they aren't bad, but they won't tell you anything important about the movement: a large tick volume may be both trending and flat. It is better to look at the increase of rates for a certain time (say, for the timeframe), it will be more informative.
Now - briefly in essence. Code the movements of pairs by one of three classes:
-1 - strong downward movement,
0 - weak movement,
+1 - strong upward movement.
Boundaries between moves are best calculated from the quantiles of the probability distribution of returns, breaking it down into three quantiles. Or just from nothing if you are too lazy.
Take into account the position of the analyzed currency pair in order to put it in the correct class. For example, if you analyse the franc, and the USDCHF pair is rising strongly (+1), the franc will fall, i.e. it will be -1 for the franc.
Next, make a vector of the state of the gas. Suppose we have nine chif pairs, we get, say, the following numbers: <+1,0,-1,0,+1,-1,0,0,+1>. This is the microstate of the gas, with the class of a particular pair at each position. All possible microstates of a gas are equally probable.
The macrostate of a gas is a set of equivalent microstates. For the macrostate of the gas in this sequence, the order of these numbers is actually unimportant. Only the quantities of different classes are important. This is where you get different probabilities. There's three +1, two -1 and four 0. Something very similar to a chiff flatt, by the way. That's what it is. A flat is usually the most likely macro state.
The thermodynamic probability of that macro-state (the number of equivalent micro-states - see underlined) is calculated through factorials: 9!/(2!*3!*4!)=1260. (By the way, the most probable macrostate is the perfect chif flatt: 9!/(3!*3!*3!)=1680. The perfect flat in this case is three +1s, three 0s and three -1s).
And, say, a microstate <+1,+1,0,+1,+1,+1,+1,+1,0,+1> corresponds to a macrostate with thermodynamic probability 9!/(7!*2!)=36. And this macro-state, as we see, is 35 times less likely than the previous macro-state. It is actually very similar to a trend.
A trend is a rare macrostate. The trend on a pair is an interesting dick, but it is only reliably registerable when multicurrency is taken into account. A trend can only be reliably registered by currency. A flat is trickier as it is more complex. Actually, there are several types of flats, but I will not go into details.
By the way, if you take the logarithm of thermodynamic probability, you get almost entropy. This entropy will probably follow me all my life: then informational (in feature selection branch), then physical :).
You can do your own digging from here. If you have any objections or interesting ideas, please write to me. Only, please, do not ask idiotic questions about what is quantile, how factorials are calculated or why probability is counted through factorials. You don't look like someone who needs to bring everything on a silver platter. Try reading this nonsense a few times and make sense of it. It took me a few months to crystallise this crap in my brain.
P.S. We could go into some very delirious stuff about phase transitions (gaseous/liquid/dead flats and crystalline trends), but that's enough for a first outline. And, of course, there are plenty of pitfalls too.
Do you mind if I join the discussion? Something is getting interesting.
By the way, yes, reducing the session time on the forum adds to the fun of the overall picture.
The currency state vector proposed by Alexey has informational entropy - assuming equal probability outcomes of its states - 14.26466 bits. But apparently, the probability density function of its states is not uniform, so the entropy will be less.
How to formalise the trend in this formulation? Sum of elements greater than zero for an uptrend and less than zero for a downtrend?
Further on, I see the following generalized scheme of actions. We calculate the probability of the trend events considering the sign, of course. And we proceed to the prediction of the currency trend.